Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/multidisciplinary-digital-publishing-institute/a-quantitative-method-for-evaluation-of-visual-privacy-in-residential-yvg7er0CT0
References
References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.
- Publisher
- Multidisciplinary Digital Publishing Institute
- Copyright
- © 1996-2021 MDPI (Basel, Switzerland) unless otherwise stated Disclaimer The statements, opinions and data contained in the journals are solely those of the individual authors and contributors and not of the publisher and the editor(s). MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Terms and Conditions Privacy Policy
- ISSN
- 2075-5309
- DOI
- 10.3390/buildings11070272
- Publisher site
- See Article on Publisher Site
Abstract
buildings Article A Quantitative Method for Evaluation of Visual Privacy in Residential Environments 1 , 2 1 , 3 , 1 2 He Zheng , Bo Wu * , Heyi Wei * , Jinbiao Yan and Jianfeng Zhu School of Geography and Environment, Jiangxi Normal University, Nanchang 330022, China; zh_factor@126.com (H.Z.); jbyan@hynu.edu.cn (J.Y.) School of Geomatics and Geoinformation, Jiangxi College of Applied Technology, Ganzhou 341000, China; 2015026023@chd.edu.cn Geodesign Research Centre, Jiangxi Normal University, Nanchang 330022, China * Correspondence: wavelet778@sohu.com (B.W.); weihy@whu.edu.cn (H.W.); Tel.: +86-791-88120251 (B.W.); +86-0791-88120430 (H.W.) Abstract: With the rapid expansion of high-rise and high-density buildings in urban areas, visual privacy has become one of the major concerns affecting human environmental quality. Evaluation of residents’ visual exposure to outsiders has attracted more attention in the past decades. This paper presents a quantitative indicator; namely, the Potential Visual Exposure Index (PVEI), to assess visual privacy by introducing the damage of potential visual incursion from public spaces and neighbor- hoods in high-density residences. The method for computing the PVEI mainly consists of three steps: extracting targets and potential observers in a built environment, conducting intervisibility analysis and identifying visible sightlines, and integrating sightlines from building level and ground level to compute the PVEI value of each building opening. To validate the proposed PVEI, a case study with a sample building located at the center of Kowloon, Hong Kong, was evaluated. The results were in accordance with the common-sense notion that lower floors are subjected to poor visual privacy, Citation: Zheng, H.; Wu, B.; Wei, H.; and privacy is relatively well-preserved in upper floors in a building. However, residents of middle Yan, J.; Zhu, J. A Quantitative Method floors may suffer the worst circumstances with respect to visual privacy. The PVEI can be a useful for Evaluation of Visual Privacy in indicator to assess visual privacy and can provide valuable information in architectural design, hotel Residential Environments. Buildings room selection, and building management. 2021, 11, 272. https://doi.org/ 10.3390/buildings11070272 Keywords: visual privacy; visual exposure; Potential Visual Exposure Index (PVEI); quantitative tools; assessment Academic Editor: Alessandro Cannavale Received: 7 June 2021 Accepted: 24 June 2021 1. Introduction Published: 26 June 2021 In the visual perception of residents, environmental quality is not only determined by what one can see, but also by the level that is visually exposed to others. Numerous studies Publisher’s Note: MDPI stays neutral focus on the evaluation of residents’ visual accessibility to landscapes [1,2], or a specified with regard to jurisdictional claims in landscape such as a green area [3–7], a water body [8–11], or a park [12,13]. Compared to published maps and institutional affil- the research on “looking out”, the issues caused by “strangers’ views in”, which refers to iations. residents’ visual exposure, have not been adequately addressed yet in the past decades, especially regarding the aspect of privacy. As one of the primary human requirements [14], privacy is a basic right of every person, and should be protected in every society by laws that guarantee this right, especially Copyright: © 2021 by the authors. at one’s residence [15]. When people choose an environment for living, they choose not Licensee MDPI, Basel, Switzerland. only the construction quality or interior-design style, but also the particular environmental This article is an open access article quality of the building [16]. However, as far as visual perception is concerned, visual distributed under the terms and exposure, which refers to privacy, has become one of the main issues that affect the conditions of the Creative Commons satisfaction of residents and the attractiveness of a built environment [17]. Visual privacy Attribution (CC BY) license (https:// is related to visual penetration between public and private domains, and deals with the creativecommons.org/licenses/by/ visibility of outsiders to residents in the built environment [14]. It is defined as the ability to 4.0/). Buildings 2021, 11, 272. https://doi.org/10.3390/buildings11070272 https://www.mdpi.com/journal/buildings Buildings 2021, 11, 272 2 of 19 carry out daily activities at home without being observed by outsiders, including neighbors and passers-by [18,19]. Previous studies have explored related elements of visual privacy from different perspectives, including architectural design of interior [20–24] and exterior [25–27] domains, and residential satisfaction [28–30]. However, the main purpose of these studies was to investigate how and to what extent specific elements affect visual privacy with qualitative methods, which failed to give a numerical result to show how much visual privacy that a particular dwelling represented exactly. Moreover, the qualitative method produced subjective preferences that resulted in the uncertainty of the level of privacy due to the limited number of participants. Therefore, understanding and evaluating visual privacy has become a crucial and urgent task in resolving conflicts and improving environmental quality in urban life [31]. The widely used method of isovist [32,33] has been adopted to quantitatively evaluate residential visual privacy in previous works [31,34,35]. For instance, Lonergan et al. [35] has proposed a quantitative approach to analyze the visibility between two sample buildings to examine residential visual privacy. He first created a set of isovists from lines along window facades of each floor, and then used a classification scheme to represent cumulative visibility and privacy from floor to floor. The application of 3D isovists provided a possibility to quantitatively study urban privacy, but this work was limited to giving a detailed solution for the understanding of urban privacy. In addition, Hwang and Lee [31] proposed an interesting method to simulate the rays from the ground level that penetrated through the opening into a building for the quantification of indoor exposure spaces. This work introduced the refuge area ratio (RAR) for the representation of visual privacy along with a method of visual simulation and numerical measurement. The refuge area was calculated as a space that was invisible to the observers inside the building. It introduced an algorithm to achieve a cumulative result by calculating the exposure area from sightline incursion of surroundings, regardless of the view distance or direction. Based on of the definition of visual exposure; that is, the opportunity to measure the focus of investigation and visual frequency [36], Shach-Pinsly et al. and Shach-Pinsly [17,37] reported that there was a high similarity between an understanding of visual exposure and the visual perception of privacy. This research attempted to quantify visual exposure of a building opening related to privacy aspects in the built environment, and presented an analysis model using the consideration of building level and street level, while the individual analysis was conducted separately, meaning that the method failed to provide an integrated, uniform, and full-scale result for each façade opening. Furthermore, the visual exposure level was considered to be affected only by the number and length of the dominant sightline. We argue that view distance, direction, and the area of observer and target openings both contribute to the evaluation of residential visual privacy. Inspired by the work in [17], we proposed a quantitative method toward visual exposure as an indicator to assess visual privacy in a residential environment. The proposed method provides a mathematical and systematic solution for subjective assessment of visual privacy in the residential environment, and allows us to have a deep understanding of visual privacy in our daily life. An indicator we call the Potential Visual Exposure Index (PVEI) is proposed in this study to evaluate the quality of a building opening’s visual privacy by quantifying the amount of visual penetration to the opening. The PVEI can be described as the damage of potential visual incursion from public spaces and neighborhoods. This indicator can support planners, architects and decision-makers in evaluating and implementing urban privacy issues in the development of an urban design. 2. Materials and Methods The model presented in this paper builds on the concept of visual exposure related to the private domain, establishing a visual indicator able to describe the visibility of features of interest (FOI), which refers to openings of building façades and pedestrian spaces on the ground. This differs from the 3D isovist or viewshed approach, which quantifies the Buildings 2021, 11, 272 3 of 19 2. Materials and Methods The model presented in this paper builds on the concept of visual exposure related to the private domain, establishing a visual indicator able to describe the visibility of fea- tures of interest (FOI), which refers to openings of building façades and pedestrian spaces Buildings 2021, 11, 272 3 of 19 on the ground. This differs from the 3D isovist or viewshed approach, which quantifies the space or volume around the observer, as here the attention is on how many observers can potentially view the target in the built environment. It involves a function of visibility space or volume around the observer, as here the attention is on how many observers can analysis from area to area, meaning the visual privacy of an opening can be represented potentially view the target in the built environment. It involves a function of visibility to measure the opening’s visual exposure only from observers of the building level and analysis from area to area, meaning the visual privacy of an opening can be represented ground level, instead of the whole surrounding spaces. to measure the opening’s visual exposure only from observers of the building level and ground level, instead of the whole surrounding spaces. 2.1. Definition A target-centered sphere was used to describe all potential observers 2.1. Definition around the target, as shown in Figure 1. From Target O on the façade of a building, the A target-centered sphere HH’VV’ was used to describe all potential observers around radius of the sphere refers to the maximum visual distance. Theoretically, all potential the target, as shown in Figure 1. From Target O on the façade of a building, the radius of the observers who can see the target O have an impact on the privacy of the target. It is noted sphere refers to the maximum visual distance. Theoretically, all potential observers who that observers within one hemispheroid cannot visually pass because of the obstruction can see the target O have an impact on the privacy of the target. It is noted that observers by façade O, while nonvisible observers can be identified with the intervisibility analysis. within one hemispheroid cannot visually pass because of the obstruction by façade O, while Point M travels from to in horizon, and travels from to in vertical, nonvisible observers can be identified with the intervisibility analysis. Point M travels meaning the horizontal and vertical ranges from 0 to and −/2 to /2 , respec- from H’ to H in horizon, and N’ travels from V’ to V in vertical, meaning the horizontal a tively. This figure presents the three-dimensional space in which all potential observers and vertical b ranges from 0 to and /2 to /2, respectively. This figure presents the can be located. three-dimensional space in which all potential observers can be located. Figure 1. Illustration of potential observers within a target-centered sphere. Figure 1. Illustration of potential observers within a target-centered sphere. For an opening of a building, its PVEI is defined as the visual incursion by observers from public spaces and building openings. The value of PVEI depends on the distribution of potential observers, which involves the view distance, view direction, and open spaces in which both observers and targets are located. Figure 2 illustrates the concept of the PVEI in detail. Suppose a resident O (target) is standing on the 12th floor of Building I, and observers A, B, C, and D are located at Building J with the same building façade. Resident O has the potential to be viewed by observers A, B, C, and D. The sightlines are formed by Buildings 2021, 11, 272 4 of 19 For an opening of a building, its PVEI is defined as the visual incursion by observers from public spaces and building openings. The value of PVEI depends on the distribution of potential observers, which involves the view distance, view direction, and open spaces in which both observers and targets are located. Figure 2 illustrates the concept of the Buildings 2021, 11, 272 4 of 19 PVEI in detail. Suppose a resident O (target) is standing on the 12th floor of Building I, and observers A, B, C, and D are located at Building J with the same building façade. Resident O has the potential to be viewed by observers A, B, C, and D. The sightlines are the resident O and the observers as OA, OB, OC, and OD. a and b are the horizontal and formed by the resident O and the observers as OA, OB, OC, and OD. and are the vertical angles of OD, respectively. horizontal and vertical angles of OD, respectively. Figure 2. The target and observers between two buildings. Figure 2. The target and observers between two buildings. In In t this his s study tudy,, w wee a assumed ssumed that that i iff an an o observer bserver is is able able t to o see the t see the tar arg gets, ets, the obser the observer ver simultaneously simultaneously wil willl be seen by the ta be seen by the tarrgets ( gets (it it i is s noted noted that that the a the assumption ssumption i is s not perf not perfectly ectly true in a particular situation). As such, view distance from an observer to the target plays true in a particular situation). As such, view distance from an observer to the target plays a a g gr reat eat role role in the determination o in the determination of f the the PVEI PVEI va value, lue, asas previo previous us studie studies s hahave ve real realized. ized. In In addition, sightlines with different directions can obviously contribute to PVEI value, as addition, sightlines with different directions can obviously contribute to PVEI value, as shown in Figure 2; sightline OA and OD with the same length (view distance; and different shown in Figure 2; sightline OA and OD with the same length (view distance; and differ- directions either in the horizontal or vertical domains. ent directions either in the horizontal or vertical domains. Similarly, the area of openings also affects the PVEI. The larger area of a target means Similarly, the area of openings also affects the PVEI. The larger area of a target means a higher possibility to be visually incursive and results in higher PVEI, while the larger a higher possibility to be visually incursive and results in higher PVEI, while the larger area of an observer signifies a higher possibility of visual penetration into the target and area of an observer signifies a higher possibility of visual penetration into the target and contributes to a higher PVEI, and vice versa. contributes to a higher PVEI, and vice versa. So far, we have concluded that the variables including the area where the observer and So far, we have concluded that the variables including the area where the observer target are located, and visual distance and view direction between a target and observers, and target are located, and visual distance and view direction between a target and ob- have a contribution in the determination of the PVEI. Therefore, the function of the PVEI servers, have a contribution in the determination of the PVEI. Therefore, the function of can be expressed as follows: the PVEI can be expressed as follows: PV = E I = F(( d ))∗f (( a rea) ) ? ∗∅ (a () )∗ j(()b) (1) (1) It is noted that the equation can be applied to the observers from both the ground It is noted that the equation can be applied to the observers from both the ground level level and building level, as shown in Figure 3. In order to determine the PVEI of each and building level, as shown in Figure 3. In order to determine the PVEI of each opening, opening, it is essential to investigate all the elements in the mathematical function. The it is essential to investigate all the elements in the mathematical function. The subfunctions subfunctions of the area, visual distance, horizontal angle, and vertical angle are discussed of the area, visual distance, horizontal angle, and vertical angle are discussed separately in the separ following ately in tsections. he following sections. Buildings 2021, 11, 272 5 of 19 Buildings 2021, 11, 272 5 of 19 Figure 3. Sightlines from building level and ground level. Figure 3. Sightlines from building level and ground level. 2.2. Variables in Mathematical Description 2.2. Variables in Mathematical Description 2.2.1. Visual Distance and Opening Area 2.2.1. Visual Distance and Opening Area Since the visual perception of an object in space reduces with the increase of distance Since the visual perception of an object in space reduces with the increase of distance from the observer [38–40], the concept of distance decay has been widely applied in from the observer [38–40], the concept of distance decay has been widely applied in visi- visibility analysis. Kyung [41] proposed the index of area-weighted visual exposure, which bility analysis. Kyung [41] proposed the index of area-weighted visual exposure, which takes distance into account to achieve an accurate assessment of visual perception. Based takes distance into account to achieve an accurate assessment of visual perception. Based on the concept that visible objects are inversely proportional to the squared distance, the on the concept that visible objects are inversely proportional to the squared distance, the index can be mathematically expressed as V = 1/d . Similarly, the function to describe the index can be mathematically expressed as = 1/ . Similarly, the function to describe phenomenon of visibility decay [42] has been proven to be a simple and effective way to the phenomenon of visibility decay [42] has been proven to be a simple and effective way realize visual-related matters. to realize visual-related matters. Therefore, we adopted this concept to describe the PVEI in this study. The variable-view Therefore, we adopted this concept to describe the PVEI in this study. The variable- distance that contributes to the PVEI can be expressed as 1/d for F(d). While the observer is view distance that contributes to the PVEI can be expressed as 1/ for ( ). While the located at an opening that is an element of area in the reality, the sub-function thus can be observer is located at an opening that is an element of area in the reality, the sub-function updated as A /d . Similarly, the target can be considered as a façade of opening as well. obs thus can be updated as /d². Similarly, the target can be considered as a façade of open- Therefore, it was further updated, and is expressed as Equation (2) for F(d, area). ing as well. Therefore, it was further updated, and is expressed as Equation (2) for (, ). A A obs tar F(d) f (area) = (2) (2) ( ) ∗ ( ) = 2.2.2. Visual Direction Previous studies took visual distance into account for the visibility analysis, enabling 2.2.2. Visual Direction us to obtain much more detailed results for visual privacy. However, few studies considered Previous studies took visual distance into account for the visibility analysis, enabling the visual direction as a variable in the quantifying expression of the visibility phenomenon. us to obtain much more detailed results for visual privacy. However, few studies consid- Since the same visual distance from the target to observers at different locations could ered the visual direction as a variable in the quantifying expression of the visibility phe- contribute a huge difference in the risk of visual incursions, we argue that it is necessary to nomenon. Since the same visual distance from the target to observers at different locations take the visual direction into account to deal with the matter of visual exposure. could contrib Figure 4 shows ute a hu the ge di layout fference of two in the risk o buildings f vis X and ual incurs Y viewed ionsfr , we arg om theutop. e that T it is nec- arget O is essa in building ry to take the visua X, and observers l directiA on i and nto a B ar ccoun e in t to deal building wi Y.th the m Target O atter of visua and observers l exposure. A and B are standing at the same level of elevation. Suppose the length of sightline AO is the Figure 4 shows the layout of two buildings X and Y viewed from the top. Target O same is in bu as BO: ilding AO X, and is perpendicular observers Ato an the d B façade are in buildin of building g Y. Target O X to which and observer resident O belongs, s A and and B arBO e standin appears g at the to besam an e acute level of angle elevaation. with Suppose the the façade. It length is obvious of sightli that ne AO i the resident s the is faced with a greater threat of visual incursion from observer A compared to B due to same as BO: AO is perpendicular to the façade of building X to which resident O be- Buildings 2021, 11, 272 6 of 19 Buildings 2021, 11, 272 6 of 19 longs, and BO appears to be an acute angle with the façade. It is obvious that the resi- dent is faced with a greater threat of visual incursion from observer A compared to B the higher level of visual penetration from location A. In this case, the angle a formed by due to the higher level of visual penetration from location A. In this case, the angle visible sightline and the resident’s façade travels from 0 to , the most risk with respect to formed by visible sightline and the resident’s façade travels from 0 to , the most risk visual exposure occurs when the sightline travels to AO, and the level of visual exposure with respect to visual exposure occurs when the sightline travels to AO, and the level of gradually decreases on both sides of AO. visual exposure gradually decreases on both sides of AO. Figure 4. Horizontal domain (top view). Figure 4. Horizontal domain (top view). Therefore, a function of sine with the variable of horizontal angle can be used to Therefore, a function of sine with the variable of horizontal angle can be used to describe the phenomenon (Function 3) reasonably. Specifically, represents the angle describe the phenomenon (Function 3) reasonably. Specifically, represents the angle formed by the sightline and the residential façade, and it can be calculated simply by a formed by the sightline and the residential façade, and it can be calculated simply by a subtraction from the azimuth of residential façade to the sightline, with the absolute value subtraction from the azimuth of residential façade to the sightline, with the absolute value granted to if the calculation results appear to be negative. Apparently, the value of granted to if the calculation results appear to be negative. Apparently, the value of ranges from 0 to due to the insignificant form of the invisible sightline when is situ- ranges from 0 to due to the insignificant form of the invisible sightline when is situated ated from to 2 . Thus, the effect of visual exposure from the horizontal domain of the from to 2. Thus, the effect of visual exposure from the horizontal domain of the visual visual direction can be quantified with the use of Function 3: direction can be quantified with the use of Function 3: ( ) ∅ = (3) ?(a) = sin a (3) Assume there is an observation point E on the upper floor (as shown in Figure 5). Assume there is an observation point E on the upper floor (as shown in Figure 5). Every opening of the observer is regarded as a single point to be analyzed in comparison. Every opening of the observer is regarded as a single point to be analyzed in comparison. The red line represents the line of sight from each observer penetrating into the resident’s The red line represents the line of sight from each observer penetrating into the resident’s room. Figure 5a–c show three types of “visible corridors” (green area) with different areas. room. Figure 5a–c shows three types of “visible corridors” (green area) with different Point C has the visible corridor with the largest volume, which means C is the most pen- areas. Point C has the visible corridor with the largest volume, which means C is the etrating. most penetrating. Figure 5. Visual penetration into rooms in different cases: (a) upper floor; (b) lower floor; (c) same floor level. Buildings 2021, 11, 272 6 of 19 longs, and BO appears to be an acute angle with the façade. It is obvious that the resi- dent is faced with a greater threat of visual incursion from observer A compared to B due to the higher level of visual penetration from location A. In this case, the angle formed by visible sightline and the resident’s façade travels from 0 to , the most risk with respect to visual exposure occurs when the sightline travels to AO, and the level of visual exposure gradually decreases on both sides of AO. Figure 4. Horizontal domain (top view). Therefore, a function of sine with the variable of horizontal angle can be used to describe the phenomenon (Function 3) reasonably. Specifically, represents the angle formed by the sightline and the residential façade, and it can be calculated simply by a subtraction from the azimuth of residential façade to the sightline, with the absolute value granted to if the calculation results appear to be negative. Apparently, the value of ranges from 0 to due to the insignificant form of the invisible sightline when is situ- ated from to 2 . Thus, the effect of visual exposure from the horizontal domain of the visual direction can be quantified with the use of Function 3: ∅( ) = (3) Assume there is an observation point E on the upper floor (as shown in Figure 5). Every opening of the observer is regarded as a single point to be analyzed in comparison. The red line represents the line of sight from each observer penetrating into the resident’s Buildings 2021, 11, 272 7 of 19 room. Figure 5a–c show three types of “visible corridors” (green area) with different areas. Point C has the visible corridor with the largest volume, which means C is the most pen- etrating. Buildings 2021, 11, 272 7 of 19 Figure 5. Visual penetration into rooms in different cases: (a) upper floor; (b) lower floor; (c) same floor level. Figure 5. Visual penetration into rooms in different cases: (a) upper floor; (b) lower floor; (c) same floor level. Similarly, in the vertical domain (Figure 6), with the same visual distance, observer Similarly, in the vertical domain (Figure 6), with the same visual distance, observer C brings a much higher threat of visual penetration to target O compared to observer D. C brings a much higher threat of visual penetration to target O compared to observer D. This is caused by the change of visual direction in vertical domain. This is caused by the change of visual direction in vertical domain. Figure 6. Vertical domain (side view). Figure 6. Vertical domain (side view). As shown in Figure 6, when the sightline from OD travels to OC, approaches 0, As shown in Figure 6, when the sightline from OD travels to OC, b approaches 0, the the PVEI reaches the maximum value, and location C is considered as the highest risk PVEI reaches the maximum value, and location C is considered as the highest risk with with respect to visual exposure of target O. As the sightline keeps traveling, the visual respect to visual exposure of target O. As the sightline keeps traveling, the visual exposure exposure gradually decreases. Therefore, a function of cosine with the variable of vertical gradually decreases. Therefore, a function of cosine with the variable of vertical angle b angle was used to describe the visual impact from vertical domain (Function (4)), and was used to describe the visual impact from vertical domain (Function (4)), and the value of the value o b ranges f fr om range /2 s from to /2: −/2 to /2 : j(b) = cos b (4) () = (4) 2.3. Model Equation As shown in Section 2.2, it is possible to provide a quantitative solution to derive 2.3. Model Equation the PVEI and contribute to the estimation of visual privacy in residential environments. As shown in Section 2.2, it is possible to provide a quantitative solution to derive the By integrating Equations (2)–(4), the PVEI of a certain opening can be quantified as: PVEI and contribute to the estimation of visual privacy in residential environments. By integrating Equations (2)–(4), the PVEI of a certain opening can be quantified as: A A i j PV E I = sin a cos b (5) i j = × × (5) In addition, considering the sum of observers from both the ground level and building In addition, considering the sum of observers from both the ground level and build- level in the residential environment, the PVEI of each opening can thus be defined as: ing level in the residential environment, the PVEI of each opening can thus be defined as: A A i j PV E I = sin a cos b (6) i å i j i j = × × (6) j i j where : visual exposure of the ℎ openness; : area of the ith target surface; : area of the ℎ surface; and : distance from the center of ℎ target to the center of ℎ observer. 3. Case Study 3.1. Description Hong Kong is regarded as one of the most densely populated cities in the world. It has more than seven million people inhabiting 1068 km of land, and the population den- sity increased from 6352 persons per km in 2006 to 6777 in 2016 [43]. As a result, the physical density of Hong Kong, as a universal geographic and spatial concern, has long been a hotspot in research [44]. Kowloon is the peninsula to the northern part of Hong Buildings 2021, 11, 272 8 of 19 where PVEI : visual exposure of the ith openness; A : area of the ith target surface; A : area i i i of the jth surface; and d : distance from the center of ith target to the center of jth observer. ij 3. Case Study 3.1. Description Hong Kong is regarded as one of the most densely populated cities in the world. It has more than seven million people inhabiting 1068 km of land, and the population Buildings 2021, 11, 272 8 of 19 density increased from 6352 persons per km in 2006 to 6777 in 2016 [43]. As a result, the physical density of Hong Kong, as a universal geographic and spatial concern, has long been a hotspot in research [44]. Kowloon is the peninsula to the northern part of Kong Island, with over 2.2 million people living in an area of less than 47 km [43]. Met- Hong Kong Island, with over 2.2 million people living in an area of less than 47 km [43]. ropa Metr rk Hotel opark Hotel Kowl Kowloon oon (Figure (Figur 7), elo 7ca ), located ted at the centra at the cent l hea ral r heart t of Kowl of Kowloon, oon, with with 75.9 75.9 m of m ab ofsol absolute ute elev elevation ation and 1 and 6 16 typ typical ical floor floors, s, is is rer presented as a typi epresented as a typical cal bui building lding to ref to reflect lect the the loca local l env envir iron onment, ment, and and wa was s tthus hus se selected lected a as s tthe he s sample ample bu building ilding in in tthis his st study udy.. Figure 7. Location of Metropark Hotel Kowloon, Kowloon, Hong Kong, China. Figure 7. Location of Metropark Hotel Kowloon, Kowloon, Hong Kong, China. 3.2. Preconditions 3.2. Preconditions In this research, three core sets of data, including the 3D Photo-Realistic Model (Plan- In this research, three core sets of data, including the 3D Photo-Realistic Model (Plan- ning Department of Hong Kong) in OSGB format, the 3D Pedestrian Network (Hong ning Department of Hong Kong) in OSGB format, the 3D Pedestrian Network (Hong Kong Geodata Store) in GDB format, and the iB1000 product of the Topographic Map Kong Geodata Store) in GDB format, and the iB1000 product of the Topographic Map (Hong Kong Geodata Store) in DWG format, could all be downloaded freely online, and (Hong Kong Geodata Store) in DWG format, could all be downloaded freely online, and were collected for preprocessing. were collected for preprocessing. A few preconditions have to be set up for the purpose of computational simplifica- A few preconditions have to be set up for the purpose of computational simplification. tion. First, viewpoints from the pedestrian were assumed to be 1.5 m off toward the First, viewpoints from the pedestrian were assumed to be 1.5 m off toward the ground. ground. Second, both building façade and ground surface were divided with 2 m × 2 m Second, both building façade and ground surface were divided with 2 m 2 m regular regular grids to extract potential viewpoints. Thus, every grid can be represented as a grids to extract potential viewpoints. Thus, every grid can be represented as a viewpoint viewpoint with the property of unified area (4 m²) in order to perform the visibility anal- with the property of unified area (4 m ) in order to perform the visibility analysis. Lastly, ysis. Lastly, terrain, infrastructure, and vegetation can be potential obstacles in the visibil- ity analysis, but they were not included in this study for the sake of simplifying the com- putation. 3.3. Preprocessing The 3D Photo-Realistic Model was formulated based on aerial photos captured from different points of view; it is a high-quality texture surface model in three dimensions that shows external features of buildings, trees, infrastructure, and terrain. The data set was used to extract all the façade openings of the study area, including the sample building (Figure 8a). Every opening was divided into regular grids as mentioned previously, and each grid was represented as a viewpoint located at the center of the grid. The 3D Pedes- Buildings 2021, 11, 272 9 of 19 terrain, infrastructure, and vegetation can be potential obstacles in the visibility analysis, but they were not included in this study for the sake of simplifying the computation. 3.3. Preprocessing The 3D Photo-Realistic Model was formulated based on aerial photos captured from different points of view; it is a high-quality texture surface model in three dimensions that shows external features of buildings, trees, infrastructure, and terrain. The data set was used to extract all the façade openings of the study area, including the sample building (Figure 8a). Every opening was divided into regular grids as mentioned previously, and each grid was represented as a viewpoint located at the center of the grid. The 3D Pedestrian Network aims at improving the walkability and connectivity of outdoor civil activities. Buildings 2021, 11, 272 In this study, it was used to identify the potential observe 9 o rs f 19 of the passers-by. The 2D building footprint of the study area was extracted from the iB1000 product, and further converted to 2.5D models as the visibility obstacles in multipatch format based on the converted to 2.5D models as the visibility obstacles in multipatch format based on the elevation attribute. The integration of two types of the data set was preprocessed, and this elevation attribute. The integration of two types of the data set was preprocessed, and this is shown in Figure 8b. is shown in Figure 8b. (a) (b) Figure 8. Data sets of the 3D Photo-Realistic Model, 3D Pedestrian Network, and 2D building foot- Figure 8. Data sets of the 3D Photo-Realistic Model, 3D Pedestrian Network, and 2D building footprint in the study area. (a) print in the study area. (a) 3D Photo-Realistic Model. (b) 3D Pedestrian Network and 2.5D Model. 3D Photo-Realistic Model. (b) 3D Pedestrian Network and 2.5D Model. 3.4. Procedure for Deriving the Potential Visual Exposure Index A quantitative assessment of visual privacy was interpreted on the basis of a sample building. The following section details the procedures outlined in Figure 9. The operation was conducted with off-the-shelf functionalities in ArcGIS Pro 2.5, Esri, Redlands, CA, USA. Buildings 2021, 11, 272 10 of 19 3.4. Procedure for Deriving the Potential Visual Exposure Index A quantitative assessment of visual privacy was interpreted on the basis of a sample Buildings 2021, 11, 272 10 of 19 building. The following section details the procedures outlined in Figure 9. The operation was conducted with off-the-shelf functionalities in ArcGIS Pro 2.5, Esri, Redlands, CA, USA. Figure 9. The implementation procedure of the PVEI estimation. Figure 9. The implementation procedure of the PVEI estimation. Two functionalities were used for the derivation of PVEI, including Construct sight Two functionalities were used for the derivation of PVEI, including Construct sight lines and Intervisibility. The Construct sight lines function generates line elements that lines and Intervisibility. The Construct sight lines function generates line elements that represent the light of sight from observer points to target features, the visibility of sight- represent the light of sight from observer points to target features, the visibility of sightlines lines is determined by the function of Intervisibility based on potential obstacles defined is determined by the function of Intervisibility based on potential obstacles defined by by the combination of 3D elements and surfaces. First, we created sightlines using ob- the combination of 3D elements and surfaces. First, we created sightlines using observer server points and target points identified through 3D model. Target points were extracted points and target points identified through 3D model. Target points were extracted from from the sample building, resulting in 528 point-features in total. Observer points were the extr sample acted frbuilding, om other bu resulting ildings ain nd pe 528de point-featur strian routes. Thus, the potenti es in total. Observer al v points isual penetra were extracted - tions could be created with the combination of observer points from other buildings from other buildings and pedestrian routes. Thus, the potential visual penetrations could be (building level) and pedestrian routes (ground level). Next, the sightline intervisibility created with the combination of observer points from other buildings (building level) and was realized according to the constructed sightlines with the visibility obstacles (i.e., 2.5D pedestrian routes (ground level). Next, the sightline intervisibility was realized according models). With the help of the Intervisibility function in ArcGIS Pro, all the visible sight- to the constructed sightlines with the visibility obstacles (i.e., 2.5D models). With the help lines were then extracted with attributes of azimuth, vertical angle, and length at both the of the Intervisibility function in ArcGIS Pro, all the visible sightlines were then extracted building and pedestrian levels, as shown in Figure 10. Buildings 2021, 11, 272 11 of 19 with attributes of azimuth, vertical angle, and length at both the building and pedestrian levels, as shown in Figure 10. Figure 10. An example of visible sightlines between a target and observers. Figure 10. An example of visible sightlines between a target and observers. It wa It was s noted noted thathat t the genera the generating ting azimut azimuth h of the si ofght the lin sightline e was aligwas ned wi aligned th globa with l global coordinates. Since the parameter in the proposed function represents a relative value, coordinates. Since the parameter a in the proposed function represents a relative value, it can be calculated with a further computation. Each orientation of the sample building it can be calculated with a further computation. Each orientation of the sample building façade is presented in Figure 11; the red dots along the building’s edge indicate the loca- façade is presented in Figure 11; the red dots along the building’s edge indicate the locations tions of façade openings, and there are a total of 33 openings in columns and 16 openings in the floor dimension aligned in all the building façades. Therefore, the value of can be calculated and then given to the corresponding visible sightline. Figure 11. Building outlines with openings on the façades and façade orientations. Other elements, including vertical angle and visual distance (3D length of the sight- line), can be created and directly imported together with into the mathematical model for the final computation. Finally, the PVEIs of the sample building were calculated using the proposed model. Buildings 2021, 11, 272 11 of 19 Figure 10. An example of visible sightlines between a target and observers. It was noted that the generating azimuth of the sightline was aligned with global coordinates. Since the parameter in the proposed function represents a relative value, Buildings 2021, 11, 272 11 of 19 it can be calculated with a further computation. Each orientation of the sample building façade is presented in Figure 11; the red dots along the building’s edge indicate the loca- tions of façade openings, and there are a total of 33 openings in columns and 16 openings in the floor dimension aligned in all the building façades. Therefore, the value of can of façade openings, and there are a total of 33 openings in columns and 16 openings in be calculated and then given to the corresponding visible sightline. the floor dimension aligned in all the building façades. Therefore, the value of a can be calculated and then given to the corresponding visible sightline. Figure 11. Building outlines with openings on the façades and façade orientations. Figure 11. Building outlines with openings on the façades and façade orientations. Other elements, including vertical angle and visual distance (3D length of the sight- Other elements, including vertical angle and visual distance (3D length of the sight- line), can be created and directly imported together with a into the mathematical model for line), can be created and directly imported together with into the mathematical model the final computation. Finally, the PVEIs of the sample building were calculated using the for the final computation. Finally, the PVEIs of the sample building were calculated using proposed model. the proposed model. 4. Results We applied our quantitative approach to deal with various spatial datasets to compute the Potential Visual Exposure Index over a sample building. In total, 528 openings of the sample building were detected, and their indicators were calculated. A 3D-perspective result was realized with an example of D = 500 m in Figure 12a. max The PVEI varied along with the change of building orientation and floor. In addition, a two-dimensional diagram of the sample building layout is presented in Figure 12b. It can be seen that the sample building consisted of eight façades in the vertical dimension. Façades 3, 6, and 8 with no opening existed, while every point extracted from the rest of the façades represented the center of an opening, and was given a computational value of PVEI. On the same floor, openings on façades 2 and 4 were detected with relatively better privacy preservation than others, while residents located along façades 1, 5, and 7 may have to take additional actions to prevent visual penetrations from outsiders. On the façade dimension, PVEI varied along with the floor and changes among different façades. As expected, upper floors tended to have lower PVEI values, and the lowest PVEI value appeared to be on the top floor. In addition, the edge of a façade that orientated to public spaces generally appeared to have a higher PVEI value due to the wide range of visual exposure to potential outsiders. Buildings 2021, 11, 272 12 of 19 4. Results We applied our quantitative approach to deal with various spatial datasets to com- pute the Potential Visual Exposure Index over a sample building. In total, 528 openings of the sample building were detected, and their indicators were calculated. A 3D-perspective result was realized with an example of = 500 in Figure 12a. The PVEI varied along with the change of building orientation and floor. In addition, a two-dimensional diagram of the sample building layout is presented in Figure 12b. It can be seen that the sample building consisted of eight façades in the vertical dimension. Façades 3, 6, and 8 with no opening existed, while every point extracted from the rest of the façades represented the center of an opening, and was given a computational value of PVEI. On the same floor, openings on façades 2 and 4 were detected with relatively better privacy preservation than others, while residents located along façades 1, 5, and 7 may have to take additional actions to prevent visual penetrations from outsiders. On the fa- çade dimension, PVEI varied along with the floor and changes among different façades. As expected, upper floors tended to have lower PVEI values, and the lowest PVEI value appeared to be on the top floor. In addition, the edge of a façade that orientated to public Buildings 2021, 11, 272 12 of 19 spaces generally appeared to have a higher PVEI value due to the wide range of visual exposure to potential outsiders. Figure 12. A classification scheme of the visual exposure model applied to the sample building based Figure 12. A classification scheme of the visual exposure model applied to the sample building based on the PVEI of each opening: (a) 3D perspective; (b) 2D diagram from the projection of the on the PVEI of each opening: (a) 3D perspective; (b) 2D diagram from the projection of the 3D model. 3D model. Since the visual exposure of each opening was threatened by observers from two parts, the bSince the visual exposure uilding level and ground level, of each ope Figure 13 nin rg was threatened by o eveals the PVEI of all floors bservers fro with respect m two to the building level, ground level, and a combination of the two parts in the sample building. parts, the building level and ground level, Figure 13 reveals the PVEI of all floors with On the ground level, as expected, the PVEI value decreased exactly with the increase of respect to the building level, ground level, and a combination of the two parts in the sam- floor. This can be well understood with our natural perception, as it is always easier for ple building. On the ground level, as expected, the PVEI value decreased exactly with the ground observers to see the openings on the lower floors. Unlike the results for the ground increase of floor. This can be well understood with our natural perception, as it is always level, Figure 13 shows an interesting result; that is, the PVEI had the highest values in the easier for ground observers to see the openings on the lower floors. Unlike the results for middle of the floors, which was caused by the larger amount of visual incursion compared the ground level, Figure 13 shows an interesting result; that is, the PVEI had the highest to lower floors, and a stronger impact from visible sightlines compared to upper floors. values in the middle of the floors, which was caused by the larger amount of visual incur- Figure 13 presents the overall result of a combination of the two levels, showing that the sion compared to lower floors, and a stronger impact from visible sightlines compared to PVEI tended to decline with an increase in floor level. Although the impact from the upper floors. Figure 13 presents the overall result of a combination of the two levels, show- Buildings 2021, 11, 272 13 of 19 building level contributed the most to the PVEI value, the combination result appeared to ing that the PVEI tended to decline with an increase in floor level. Although the impact follow a similar trend as the results from the ground level. from the building level contributed the most to the PVEI value, the combination result appeared to follow a similar trend as the results from the ground level. Figure 13. Figure 13. PV PVEI EI of all the f of all the floors loors with respect with respect to the grou to the ground nd level, level, bu building ilding level, level, and a co and a combination mbination of the two levels. of the two levels. Similarly, we applied our method to the sample building to investigate the tendency Similarly, we applied our method to the sample building to investigate the tendency of the PVEI with respect to building facades. A total of 35 columns of the openings were of the PVEI with respect to building facades. A total of 35 columns of the openings were aligned in five facades (excluding three facades with no opening) of the sample building. Figure 14 indicates the PVEI of all the openings along the columns on the facades with respect to the ground level, building level, and a combination of the two levels in the sam- ple building. PVEI results from both the ground and building levels presented unified upward or downward trends on the same façade. However, several target points with different PVEI values located near the edge of facades were exceptions that did not follow the overall trend. For example, on the ground level, column 1 showed a great difference in PVEI values from the rest of the columns on the façade 1. Façade 1 was in close prox- imity to another building, which created a visual barrier and produced low PVEI values for column 1. Similarly, columns 19 and 20 revealed relatively lower values of PVEI on both the building and ground levels because targets on these two columns were the closest to the adjacent building. In sum, the overall tendency of the combination was similar to the building level, which indicated that the visual exposure by the observers from the building level played a significant role in the study building. Figure 14. PVEI of the building direction in columns with respect to the ground level, building level, and a combination of the two levels. Buildings 2021, 11, 272 13 of 19 Figure 13. PVEI of all the floors with respect to the ground level, building level, and a combination of the two levels. Similarly, we applied our method to the sample building to investigate the tendency of the PVEI with respect to building facades. A total of 35 columns of the openings were Buildings 2021, 11, 272 13 of 19 aligned in five facades (excluding three facades with no opening) of the sample building. Figure 14 indicates the PVEI of all the openings along the columns on the facades with respect to the ground level, building level, and a combination of the two levels in the sam- ple building. PVEI results from both the ground and building levels presented unified aligned in five facades (excluding three facades with no opening) of the sample building. upward or downward trends on the same façade. However, several target points with Figure 14 indicates the PVEI of all the openings along the columns on the facades with different PVEI values located near the edge of facades were exceptions that did not follow respect to the ground level, building level, and a combination of the two levels in the the overall trend. For example, on the ground level, column 1 showed a great difference sample building. PVEI results from both the ground and building levels presented unified in PVEI values from the rest of the columns on the façade 1. Façade 1 was in close prox- upward or downward trends on the same façade. However, several target points with imity to another building, which created a visual barrier and produced low PVEI values different PVEI values located near the edge of facades were exceptions that did not follow for column 1. Similarly, columns 19 and 20 revealed relatively lower values of PVEI on the overall trend. For example, on the ground level, column 1 showed a great difference in both the building and ground levels because targets on these two columns were the closest PVEI values from the rest of the columns on the façade 1. Façade 1 was in close proximity to the adjacent building. to another building, which created a visual barrier and produced low PVEI values for In sum, the overall tendency of the combination was similar to the building level, column 1. Similarly, columns 19 and 20 revealed relatively lower values of PVEI on both which indicated that the visual exposure by the observers from the building level played the building and ground levels because targets on these two columns were the closest to a significant role in the study building. the adjacent building. Figure 14. PVEI of the building direction in columns with respect to the ground level, building level, Figure 14. PVEI of the building direction in columns with respect to the ground level, building level, and a combination of the two levels. and a combination of the two levels. In sum, the overall tendency of the combination was similar to the building level, which indicated that the visual exposure by the observers from the building level played a significant role in the study building. Furthermore, we tested the model and applied it to the cases in which the maximum visual distance (D ) was given as 25, 50, 100, and 300 m. The various D was employed max max for the visual privacy assessment and numerical comparison. Figure 15 compares PVEI values in the floor dimension (Figure 15a) and column dimension (Figure 15b) under various D For both the floor and column dimensions, the max overall trends of PVEI in different D remained consistent. It was noted that the PVEI max value almost reached 0 as D approached 25 m on the 16th floor, which indicated there max was almost no light of sight penetrating into some of the openings on the 16th floor under the condition. Buildings 2021, 11, 272 14 of 19 Furthermore, we tested the model and applied it to the cases in which the maximum visual distance ( ) was given as 25, 50, 100, and 300 m. The various was em- ployed for the visual privacy assessment and numerical comparison. Figure 15 compares PVEI values in the floor dimension (Figure 15a) and column di- mension (Figure 15b) under various . For both the floor and column dimensions, the overall trends of PVEI in different remained consistent. It was noted that the PVEI value almost reached 0 as approached 25 m on the 16th floor, which indicated there was almost no light of sight penetrating into some of the openings on the 16th floor under the condition. In addition, the values of PVEI changed remarkably between the cases =25 and = 100 , which revealed that a large number of visible sightlines existed with lengths between 25 and 100 m. This may have been caused by the specific building layout and the density in the neighborhood of the sample building. In addition, as shown in Fig- Buildings 2021, 11, 272 14 of 19 ure 15, PVEI values did not differ greatly when the maximum visual distance was given as 100 m, 300 m, and 500 m. (a) Floor dimensions. (b) Column dimensions. Figure 15. PVEI Figure 15. values accor PVEIding valuto es ac thecord floors ing to the (a) and f the loor co s ( lumns a) and (b the ) under coluvarious mns (b) under various ranges of the ranges of the maximum visual distance. maximum visual distance. In addition, the values of PVEI changed remarkably between the cases D = 25 m max and D = 100 m, which revealed that a large number of visible sightlines existed with max lengths between 25 and 100 m. This may have been caused by the specific building layout and the density in the neighborhood of the sample building. In addition, as shown in Figure 15, PVEI values did not differ greatly when the maximum visual distance was given as 100 m, 300 m, and 500 m. In conclusion, the visual exposure of an opening may be varied dramatically among different floors and facades. On the ground level, the higher the floor, the lower the PVEI value and the better the privacy. On the building level, although there were no distinct patterns as there were for the ground level, upper floors can be a better choice to preserve residential visual privacy in the same building. In this study case, PVEI values of the sample building were mainly determined by potential observers on the building level, which can be explained by the high-rise and dense building packing in the surrounding area of the sample building. Moreover, as the D changed from 25 to 500 m, the number max of visible sightlines increased tremendously, resulting in an increase in the PVEI value and a decrease in the quality of visual privacy of each opening. We determined that it was suitable to describe the visual privacy of the sample building in the case of D = 100 m. max Overall, with the change of D , the overall trend of PVEI remained stable. max Buildings 2021, 11, 272 15 of 19 5. Discussion 5.1. Determination of the Maximum View Distance in the Model Scholars have studied visual privacy and concluded that privacy can only be invaded when the visual distance situates within a relatively small range. This is because the effect of sight incursion on visual privacy is not just about “what observers can see”, it is about the capacity of visual damage for observers to distinguish among the various forms of objects or different characteristics of people. There are a few references that discuss how the distance between buildings affects visual exposure at façade openings, and the minimum distance between buildings needed to provide sufficient visual privacy for the residents [18,45]. Mitrany [46] found that a distance of 35 m between buildings was enough to obtain the desired level of visual privacy for residents, while a distance of 10 m failed to meet the needs. Day [47] examined the street distance of low-rise neighborhoods and found that 24.4 m in distance was sufficient for the preservation of privacy. In addition, visible sightlines were categorized by visual distances with four ranges to rank the level of visual exposure of building openings, and a length of sightline greater than 50 m was considered a well-preserved level of privacy [17]. Nevertheless, in this study, we examined the visual exposure of building openings based on various ranges of the maximum view distance, including 25, 50, 100, 300 and 500 m. The results showed that when D approached 100 m, PVEI values of openings max reached a steady state. The case of a D greater than 100 m could obtain an outcome max with a better precision, and while it may be time-consuming with low effectiveness in computation, it was not sufficient to realize the potential visual exposure of the opening when the D was lower than 100 m. Consequently, the investigation with various D max max provided a possibility to identify and realize the potential sight penetrators and their spatial distributions in the built environment. Moreover, the light of sight created with the use of a telescope or other devices in a residential environment can be “shortened in length”, contributing to a much clearer scenery in the observer ’s view, and eventually changing the amount of visual exposure and the level of visual privacy. While this phenomenon was not taken into account in this, since we were mainly focused on establishing the method of the quantitative model. A forward investigation of the phenomenon in future work could be significant in a better understanding of visual privacy in residential environments. 5.2. Impact Factors on PVEI The function of the PVEI was established on a basis of visibility, and revealed a relationship among PVEI and view distance, opening area, and view direction. First, the area of target opening had a positive influence on the PVEI, and a high PVEI value increased the possibility of visual intrusion, resulting in poor visual privacy. Nowadays, the urban landscape usually has a wide-ranging and complex morphology, in which large-area openings passively accept the sight views from all directions, while visual penetrations from most of the directions are constantly blocked in the small-area openings. Second, since the PVEI is an indicator measuring each opening’s capacity of visual exposure in urban space, characteristics of buildings and the distribution of potential observers have a substantial impact on the value of PVEI. For instance, the expansion of city buildings toward density and verticality has led to a dramatic increase in the number of potential observers within a compact neighborhood at the building level, especially for large cities in which the population is concentrated in a small community. Conversely, with the increase of building spacing, the level of visual privacy that residents can preserve is more likely to have extended horizontal and vertical dimensions. Thus, buildings with more open space and lower density contribute to lower PVEI values for neighborhoods. Finally, the location and width of pedestrian routes around the residential environment have a potential effect on the PVEI, especially for occupants on low floors. Although observers from the building level had the largest impact on the sample building’s visual privacy in this study, it is important to pay attention to the damage of visual exposure from the ground level, which can be a main issue in a low-rise and spaced-out building neighborhood. Buildings 2021, 11, 272 16 of 19 from most of the directions are constantly blocked in the small-area openings. Second, since the PVEI is an indicator measuring each opening’s capacity of visual exposure in urban space, characteristics of buildings and the distribution of potential observers have a substantial impact on the value of PVEI. For instance, the expansion of city buildings toward density and verticality has led to a dramatic increase in the number of potential observers within a compact neighborhood at the building level, especially for large cities in which the population is concentrated in a small community. Conversely, with the in- crease of building spacing, the level of visual privacy that residents can preserve is more likely to have extended horizontal and vertical dimensions. Thus, buildings with more open space and lower density contribute to lower PVEI values for neighborhoods. Finally, the location and width of pedestrian routes around the residential environment have a potential effect on the PVEI, especially for occupants on low floors. Although observers from the building level had the largest impact on the sample building’s visual privacy in this study, it is important to pay attention to the damage of visual exposure from the Buildings 2021, 11, 272 16 of 19 ground level, which can be a main issue in a low-rise and spaced-out building neighbor- hood. 5.3. A Further Development Required for Deriving a Standard PVEI 5.3. A Further Development Required for Deriving a Standard PVEI Based on the mathematical function, the value of the PVEI is proportional to the area Based on the mathematical function, the value of the PVEI is proportional to the area of openings and the inverse square distance between observers and the target. However, of openings and the inverse square distance between observers and the target. However, because the compact and vertical complexes built in modern society must deal with a risk because the compact and vertical complexes built in modern society must deal with a of visual privacy in terms of the high value of the PVEI, it is not common to find a resi- risk of visual privacy in terms of the high value of the PVEI, it is not common to find a dence with the best protection of visual privacy (i.e., PVEI = 0). The PVEI value of an open- residence with the best protection of visual privacy (i.e., PVEI = 0). The PVEI value of ing approaching 0 exists in many cases; for example, an opening toward a region (e.g., sea an opening approaching 0 exists in many cases; for example, an opening toward a region or lake) without those buildings or pedestrian routes, or an opening that is totally blocked (e.g., sea or lake) without those buildings or pedestrian routes, or an opening that is totally by urban elements or trees. Visual privacy can be well preserved in these two cases, but blocked by urban elements or trees. Visual privacy can be well preserved in these two rarely occurs in reality. In addition, these openings with a low PVEI value may also be cases, but rarely occurs in reality. In addition, these openings with a low PVEI value may accompanied by environmental damage such as sunlight and ventilation. In a built envi- also be accompanied by environmental damage such as sunlight and ventilation. In a built ronment, people choose a residence with relatively better visual privacy that can meet environment, people choose a residence with relatively better visual privacy that can meet their expectations, rather than a residence with a PVEI value of 0. their expectations, rather than a residence with a PVEI value of 0. Therefore, it is of the highest importance to ensure all the openings of buildings (es- Therefore, it is of the highest importance to ensure all the openings of buildings pecially for residences and hotels) achieve an acceptable level of visual privacy, which (especially for residences and hotels) achieve an acceptable level of visual privacy, which refers to the Standard Potential Visual Exposure Index (SPVEI), which should be designed refers to the Standard Potential Visual Exposure Index (SPVEI), which should be designed by architects and urban planners. As shown in Figure 16, the value of the PVEI between 0 by architects and urban planners. As shown in Figure 16, the value of the PVEI between 0 and the SPVEI refers to a low preservation of the visual privacy of an opening, while the and the SPVEI refers to a low preservation of the visual privacy of an opening, while the value of the PVEI that is higher than the SPVEI represents an acceptable level of visual value of the PVEI that is higher than the SPVEI represents an acceptable level of visual privacy. It is worth noting that the same SPVEI value for different openings can corre- privacy. It is worth noting that the same SPVEI value for different openings can correspond spond to the different sele to the different selectionsctions of of D . Mor . M eover oreover, , SPVEI SPmay VEI may be v be varied, aried, sinc since the e the de- desired max sired visual privacy differs according to the culture, region, and functionality of buildings. visual privacy differs according to the culture, region, and functionality of buildings. Consequentl Consequently y, the proposed ma , the proposed mathematical thematical ffunction unction can be used can be usedaas s a ba a basic sic tool tool to toqquantify uantify visu visual al priv privacy acy, ,and andis h is helpful elpful tto o iidentify dentify var various ious S SPVEI PVEI under under cert certain ain circ circumstances. umstances. Figure 16. The relationship between PVEI and the selection of . Figure 16. The relationship between PVEI and the selection of D . max 6. Conclusions With the continuous improvement of income levels, urban residents are altering their priorities from basic necessities of living to the quality of their lives. Visual privacy, as a key factor in the quality of urban life, is greatly needed at every scale. Quantitative measurement and analysis of urban residents’ visual privacy or the visual penetration by strangers is an integral part of assessing the overall quality of residential life in an urban environment. In this paper, an indicator was developed to provide an objective and people-centered evaluation and quantitative analysis of the visual exposure of urban space on different openings of building façades. This indicator was calculated on the basis of a mathematical model using the data of the building footprint and the pedestrian network in a 3D perspective. In the sample building in the center of Kowloon, people who live on lower floors tended to have a high level of visual exposure or low level of visual privacy, but this did not indicate that the higher the floor, the better preservation of the visual privacy. First, at the ground level, the PVEI value of an opening consistently decreased with an increase in the building floor because of the increase of visual distance. Second, at the building Buildings 2021, 11, 272 17 of 19 level, targets on the middle floor had the highest probability to be visually exposed to the observers from all directions, and a large number of sight incursions led to the worst preservation of visual privacy. Consequently, with the integration of both levels, residents of upper floors had a relatively better preservation of their visual privacy in the building. Several possible errors that existed in the assumptions may have affected the results of the Potential Visual Exposure Index computation. For instance, the observer area and target area were independent variables in the proposed model, and the size of an opening had an impact on the calculation results for the PVEI. In the study case, although two openings could be able to form numerous sightlines in theory, only one sightline was created by extracting the center of the two openings and further introduced into the computation. However, it was not guaranteed that all sightlines between observers and targets were visible, since some parts of a target opening may have been blocked from the view of the observer. Specifically, the smaller the opening area was divided, the higher the precision of the result. The larger the opening area was divided, the lower the precision of the result. However, when increasing the size of the grids into which an opening was divided, the efficiency of the computation decreased. Therefore, it was crucial to find a trade-off between computational efficiency and precision. As an indicator capturing an important quality of residential environment, the Po- tential Visual Exposure Index in this paper can not only remind residents of the potential damage to visual privacy, but also help urban planners and architects improve the quality of urban environment by quantitatively assessing the sensory “visual exposure” value of city buildings. Author Contributions: Conceptualization, B.W.; methodology, H.Z. and H.W.; investigation, H.Z. and J.Y.; software, J.Z.; resources, B.W. and H.Z.; writing original draft preparation, H.Z.; writing review and editing, B.W., H.W. and H.Z.; visualization, J.Y. and J.Z.; supervision, B.W. and H.W.; funding acquisition, B.W. and H.W. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by Natural Science Foundation of China, grant number 41961055 and 31860233, and The National Key Research and Development Program of China, grant number 2018YFE0207800. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Publicly available data sets were analyzed in this study. The 3D Photo-Realistic Model data set can be found here: https://www.pland.gov.hk/pland_en/info_ serv/3D_models/download.htm (accessed on 10 October 2020). The 3D Pedestrian Network data set can be found here: https://geodata.gov.hk/gs/view-dataset?uuid=201eaaee-47d6-42d0 -ac81-19a430f63952&sidx=0# (accessed on 15 October 2020). The iB1000 data sets can be found here: https://www.hkmapservice.gov.hk/OneStopSystem/map-search?product=OSSCatB&series= iB1000&locale=en (accessed on 8 September 2020). Acknowledgments: The authors would like to thank the anonymous reviewers and editors for their valuable comments. The authors are also thankful to the Hong Kong Government (Planning Department and Lands Department) for providing the data sets. Conflicts of Interest: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. References 1. Arriaza, M.; Cañas-Ortega, J.F.; Cañas-Madueño, J.A.; Ruiz-Aviles, P. Assessing the visual quality of rural landscapes. Landsc. Urban Plan. 2004, 69, 115–125. [CrossRef] 2. Abd-Alhamid, F.; Kent, M.; Calautit, J.; Wu, Y. Evaluating the impact of viewing location on view perception using a virtual environment. Build. Environ. 2020, 180, 106932. [CrossRef] 3. Yang, J.; Zhao, L.; McBride, J.; Gong, P. Can you see green? Assessing the visibility of urban forests in cities. Landsc. Urban Plan. 2009, 91, 97–104. [CrossRef] Buildings 2021, 11, 272 18 of 19 4. Yu, S.; Yu, B.; Song, W.; Wu, B.; Zhou, J.; Huang, Y.; Wu, J.; Zhao, F.; Mao, W. View-based greenery: A three-dimensional assessment of city buildings’ green visibility using Floor Green View Index. Landsc. Urban Plan. 2016, 152, 13–26. [CrossRef] 5. Falfán, I.; Muñoz-Robles, C.A.; Bonilla-Moheno, M.; MacGregor-Fors, I. Can you really see ‘green’? Assessing physical and self-reported measurements of urban greenery. Urban For. Urban Green. 2018, 36, 13–21. [CrossRef] 6. Chen, J.; Zhou, C.; Li, F. Quantifying the green view indicator for assessing urban greening quality: An analysis based on Internet-crawling street view data. Ecol. Indic. 2020, 113, 106192. [CrossRef] 7. Zhang, J.; Cheng, Y.; Li, H.; Wan, Y.; Zhao, B. Deciphering the changes in residential exposure to green spaces: The case of a rapidly urbanizing metropolitan region. Build. Environ. 2021, 188, 107508. [CrossRef] 8. Bulut, Z.; Yilmaz, H. Determination of waterscape beauties through visual quality assessment method. Environ. Monit. Assess. 2009, 154, 459–468. [CrossRef] [PubMed] 9. Alphan, H.; Sonmez, F. Mapping availability of sea view for potential building development areas. Environ. Monit. Assess. 2015, 187, 413. [CrossRef] [PubMed] 10. Lin, L.; Homma, R.; Iki, K. Preferences for a lake landscape: Effects of building height and lake width. Environ. Impact Assess. Rev. 2018, 70, 22–33. [CrossRef] 11. Alphan, H.; Asur ¸ , F. Geospatial analysis of lake scenery as an indicator for the environment: The City of Van (Turkey) and its surroundings. Environ. Sustain. Indic. 2021, 9, 100091. [CrossRef] 12. Qi, T.; Zhang, G.; Wang, Y.; Liu, C.; Li, X. Research on landscape quality of country parks in Beijing as based on visual and audible senses. Urban For. Urban Green. 2017, 26, 124–138. [CrossRef] 13. Gungor, S.; Polat, A.T. Relationship between visual quality and landscape characteristics in urban parks. J. Environ. Prot. Ecol. 2018, 19, 939–948. [CrossRef] 14. Heydaripour, O.; Behmaneshnia, F.; Talebian, E.; Shahi, P.H. A survey on privacy of residential life in contemporary apartments in Iran. Int. J. Sci. Study 2017, 5, 254–263. [CrossRef] 15. Al-Birawi, S.; Adawi, A.A.D. Privacy criterions and elements in Islamic residential architecture. Jordan J. Islamic Stud. 2019, 15, 402–421. [CrossRef] 16. Rapoport, A. Culture, Architecture, and Design; Locke Science Publishing Company, Inc.: Chicago, IL, USA, 2005. 17. Shach-Pinsly, D.; Fisher-Gewirtzman, D.; Burt, M. ‘Visual exposure’ analysis model: A comparative evaluation of three case studies. Urban Design Int. 2007, 12, 155–168. [CrossRef] 18. Al-Kodmany, K. Residential visual privacy: Traditional and modern architecture and urban design. J. Urban Des. 1999, 4, 283–311. [CrossRef] 19. Tomah, A.N. Visual privacy in residential areas: Amendment of building regulations. Proc. Inst. Civ. Eng. Urban Des. Plan. 2012, 165, 43–53. [CrossRef] 20. Hashim, A.H.; Rahim, Z.A.; Rashid, S.N.S.A.; Yahaya, N. Visual privacy and family intimacy: A case study of malay inhabitants living in two-storey low-cost terrace housing. Environ. Plan. B Plan. Des. 2016, 33, 301–318. [CrossRef] 21. Georgiou, M. Architectural Privacy: A Topological Approach to Relational Design Problems. Master ’s Thesis, Bartlett School of Graduate Studies, University College London, London, UK, 2006. 22. Indraprastha, A.; Shinozaki, M. Computational models for measuring spatial quality of interior design in virtual environment. Build. Environ. 2012, 49, 67–85. [CrossRef] 23. Othman, Z.; Buys, L.; Aird, R. Observing privacy, modesty and hospitality in the home domain: Three case studies of Muslim homes in Brisbane, Australia. Int. J. Archit. Res. 2014, 8, 266–283. [CrossRef] 24. Shepperson, M. Visibility, privacy, and missing windows: Lighting domestic space in ancient Mesopotamia. In The Oxford Handbook of Light in Archaeology; Papadopoulos, C., Moyes, H., Eds.; Oxford University Press: Oxford, UK, 2017. 25. Sipahi, A. Window-conflicts in the Ottoman Empire and Turkey: Visual privacy, materiality and right to the city. Middle East. Stud. 2016, 52, 588–604. [CrossRef] 26. Hwang, J.H.; Lee, H. Parametric model for window design based on prospect-refuge measurement in residential environment. Sustainability 2018, 10, 3888. [CrossRef] 27. Alelwani, R.; Ahmad, M.W.; Rezgui, Y. Public perception of vernacular architecture in the Arabian Peninsula: The case of rawshan. Buildings 2020, 10, 151. [CrossRef] 28. Liu, A.M.M. Residential satisfaction in housing estates: A Hong Kong perspective. Autom. Constr. 1999, 8, 511–524. [CrossRef] 29. Kim, J.; De Dear, R. Workspace satisfaction: The privacy-communication trade-off in open-plan offices. J. Environ. Psychol. 2013, 36, 18–26. [CrossRef] 30. Ibem, E.O.; Opoko, A.P.; Adeboye, A.B.; Amole, D. Performance evaluation of residential buildings in public housing estates in Ogun State, Nigeria: Users’ satisfaction perspective. Front. Archit. Res. 2013, 2, 178–190. [CrossRef] 31. Hwang, J.Y.; Lee, H. 3D visual simulation and numerical measurement of privacy in traditional Korean palace. In Proceedings of the 22nd International Conference of the Association for Computer-Aided Architectural Design Research in Asia (CAADRIA), Hong Kong, China, 5–8 April 2017. 32. Benedikt, M.L. To take hold of space: Isovists and isovist fields. Environ. Plan. B Plan. Des. 1979, 6, 47–65. [CrossRef] 33. Morello, E.; Ratti, C. A digital image of the city: 3D isovists in Lynch’s urban analysis. Environ. Plan. B Plan. Des. 2009, 36, 837–853. [CrossRef] Buildings 2021, 11, 272 19 of 19 34. Alkhazmi, H.M.; Esin, N. Investigating the visual privacy on houses layouts in traditional desert settlement of Ghadames city-Libya-by using space syntax analysis. Int. J. Appl. Eng. Res. 2017, 12, 8941–8951. 35. Lonergan, C.; Hedley, N. Unpacking isovists: A framework for 3D spatial visibility analysis. Cartogr. Geogr. Inf. Sci. 2015, 43, 87–102. [CrossRef] 36. Llobera, M. Extending GIS-based visual analysis: The concept of visualscapes. Int. J. Geogr. Inf. Sci. 2003, 17, 25–48. [CrossRef] 37. Shach-Pinsly, D. Visual openness and visual exposure analysis models used as evaluation tools during the urban design development process. J. Urban. 2010, 3, 161–184. [CrossRef] 38. Kumsap, C.; Borne, F.; Moss, D. The technique of distance decayed visibility for forest landscape visualization. Int. J. Geogr. Inf. Sci. 2005, 19, 723–744. [CrossRef] 39. Taylor, P.J. Distance decay in spatial interactions. Concepts Tech. Mod. Geogr. 1975, 2, 35. 40. Chen, Y. The distance-decay function of geographical gravity model: Power law or exponential law? Chaos Solitons Fractals 2015, 77, 174–189. [CrossRef] 41. An, K.; Ko, H.; Kim, C. Development and implementation of visual exposure indicator for residential development. Int. J. Smart Home 2015, 9, 1–10. [CrossRef] 42. Tara, A.; Lawson, G.; Renata, A. Measuring magnitude of change by high-rise buildings in visual amenity conflicts in Brisbane. Landsc. Urban Plan. 2021, 205, 103930. [CrossRef] 43. Census and Statistics Department. Demographic Trends in Hong Kong 1986–2016. Statistical Reports; 2017. Available online: https://www.censtatd.gov.hk/en/EIndexbySubject.html?pcode=B1120017&scode=150 (accessed on 7 May 2021). 44. Tang, W.S.; Lee, J.W.Y.; Hui, T.W.; Yip, M.K.C. The “Urban density” question in Hong Kong: From absolute space to social processes. City Cult. Soc. 2019, 17, 46–53. [CrossRef] 45. Merry, S.E. Crowding, conflict, and neighborhood regulation. In Human Behavior and Environment; Springer: Boston, MA, USA, 1987; Volume 9, pp. 35–68. 46. Mitrany, M. Subjective Housing Density and the Housing Context. Ph.D. Thesis, Department of Architecture and Town Planning, The Technion-IIT, Haifa, Israel, 2001. 47. Day, L.L. Choosing a house: The relationship between dwelling type, perception of privacy and residential satisfaction. J. Plan. Educ. Res. 2016, 19, 265–275. [CrossRef]
Journal
Buildings – Multidisciplinary Digital Publishing Institute
Published: Jun 26, 2021
Keywords: visual privacy; visual exposure; Potential Visual Exposure Index (PVEI); quantitative tools; assessment