Access the full text.
Sign up today, get DeepDyve free for 14 days.
Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 − 75 Original paper DOI: 10.2478/agri-2021-0006 RAPID IDENTIFICATION OF RICE MACRONUTRIENT CONTENT IN SALINE SOILS USING SMARTPHONE CAMERA ADITYA NUGRAHA PUTRA, ALBERTH FERNANDO SITORUS, QUID LUQMANUL HAKIM, MARTIANA ADELYANTI, ISTIKA NITA AND SUDARTO University of Brawijaya, Malang, Indonesia Putra, A.N., Sitorus, A.F., Hakim, Q.L., Adelyanti, M., Nita, I. and Sudarto (2021). Rapid identification of rice macronutrient content in saline soils using smartphone camera. Agriculture (Poľnohospodárstvo), 67(2), 61 – 75. Indonesia’s rice production has decreased by 6.83% (on average) in the last five years (2015 – 2019) because of some factors. Salinity (42%) is one of the leading factors that cause decreasing rice production besides climate change (21%), drought (9%), and other factors (28%). The smartphone camera serves as an alternative technology to prevent macronutrient deficiencies due to salinity. This study used aerial photos from android with visible light (R, G, and B), and the image was taken from a height of 5 m. The observation of macronutrient content in plant biomass was carried out using a free grid to adjust rice fields and saline soil. The formula was obtained from regression analysis and paired t-test between the biomass macronutrient and the extracted digital number of aerial photographs that have been stacked. The results showed that digital number (DN) from a smartphone was reliable to predict nitrogen (N), phosphorus (P), and potassium (K) content in rice with formula 2 2 2 N = 0.0035 * DN + 0.8192 (R 0.84), P = 0.0049 * DN – 0.2042 (R 0.70), and K = 0.0478 * DN – 2.6717 (R 0.70). There was no difference between the macronutrient estimation results from the formula and the field’s original data. Key words: remote sensing, visible light, android, nitrogen, phosphorus, potassium, salinity Rice is one of the world’s vital food commodities search location, namely the Demak Regency, which in which productivity is predicted to decline in 2020 resulted in crop failure of 0.629% in 2013, 8.121% by 0.60% (USDA 2020). Based on the (World Ag- in 2014, and 9.173% in 2015. Another problem that ricultural Production.com, 2020), Indonesia ranked has a similar effect to salinity is climate change. third in rice producers, with production reaching Rice production has decreased due to climate change 36.5 million metric tons. In Indonesia, rice is an es- in Indramayu Regency (Ruminta 2016), with an sential food commodity and is the primary foodstuff average decline of 21% to 40%. Rice is often affect- for the community. The increase in rice demand is ed by salinity, reducing 42% production (Ahmed & not matched by increased production, which fell by Haider 2014). Rice can adapt to almost any environ- –6.83% (on average) from 2015 to 2019 (Central ment from lowland to highland. In Indonesia, rice Bureau of Statistics 2020). cultivation is carried out in various lands, including Compared to other problems that cause fluctu- wetlands in lowland rice fields, dry land, upland ations in rice production, salinity still has a more rice fields, and peatlands (Utama 2015). A study by severe impact – another case related to water avail- Mardiansyah et al. (2018) stated that the Ciherang ability. A study by (Iswari et al. 2016) mentioned variety has moderate salinity tolerant characters. that rice production is caused by drought at the re- The Inpari 32 variety is an inbred from the selec- Aditya Nugraha Putra (*Corresponding author), Soil Science Department, Faculty of Agriculture, University of Brawijaya, Malang, Indonesia. E-mail: firstname.lastname@example.org © 2021 Authors. This is an open access article licensed under the Creative Commons Attribution-NonComercial-NoDerivs License (http://creativecommons.org/licenses/by-nc-nd/4.0/). 61 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 − 75 tion results of the Ciherang variety, and Inpari 42 the colour of rice leaves captured via an android is a salinity tolerant variety (Agricultural Research camera. Image processing is a feature extraction of and Development Agency, 2019). red, green, and blue (RGB) values to obtain features Soil salinity shows the concentration of dis- in leaf colour images. From the accuracy testing solved salt in the soil (Sembiring & Gani 2010). Sa- results, the application accuracy rate in analyzing linity occurs due to (1) the high intake of water con- and recommending nitrogen needs on average is taining salt, for example, due to seawater intrusion, 66.67%. (2) higher evaporation and evapotranspiration than This study aims to implement an android cam- precipitation (rainfall), and (3) soil parent material era to monitor macronutrient content in saline fields. containing salt deposits (Rachman et al. 2018). The salinity level in a plot/landscape is considered Lowland rice production due to salinity stress can a common problem. The previous study conducted cause a decrease in production. The effect of salini- smartphone camera use (Astika et al. 2011), mac- ty on crops includes osmotic pressure, nutrient bal- ronutrient analysis, and salinity analysis (Grattan ance, and NaCl salts’ toxic impact on saline soils & Grieve 1998). It is necessary to have a techno- that can disrupt the nutrient balance because certain logical breakthrough that can "photograph" the nutrients are excess or reduced. Potassium is ex- variability macronutrients in the salinity area. The changeable, which means a decrease in these ele- breakthrough is the use of terrestrial cameras. The ments’ availability affects other nutrients crops (Se- implementation of agricultural precision is planned tiawan & Herdianto 2018). The salinity symptoms for modern technology by utilizing industrial era 4.0 in rice crops begin with dry leaf tips, reduced tillers, technologies such as terrestrial cameras. Therefore, root length, crop height, shoot dry weight, and root this study measures how accurate terrestrial cam- weight. Salinity suppresses crop growth processes eras are in analyzing salinity problems in rice crops. with effects that inhibit cell enlargement and divi- Hopefully, this study can support the government's sion, protein production, and the addition of crop food security program (NAWACITA) and SDGs. biomass. Crops that experience salt stress do not re- spond directly to damage, but growth is depressed and changes slowly. Excess Na in crop cells di- MATERIAL AND METHODS rectly damages membrane systems and organelles, causing abnormal growth and development before Research location crop death (Sayed & Sayed 2013). The research activity was carried out in the The use of smartphones to identify the macro- rice fields in the Jabon district, Sidoarjo Regency, nutrient content of biomass is interesting to study. and East Java (Figure 1). Jabon district is located This technology has potential because many people in a lowland, with coordinates of 112° 70’ 36.17” have used smartphones, have high resolutions, and – 112° 87’33.13” East longitude and 7° 49” 40.01” can be used quickly. One of the most widely used – 7° 57’ 83.45” South latitude. The study area has smartphone platforms is an android (>80% smart- an annual rainfall ranging from 1,300 – 1,700 mm phone user use). Android is a Linux-based operating per year, with the number of rainy days ranging system for smartphones that includes an operating from 80 – 120 rainy days per year. The average system, middleware, and applications (Safaat 2011). air temperature per year in this area ranges from Some android cameras’ sensors are ambient light 21 – 34°C with a relative humidity level of ± 76% sensors, temperature and humidity sensors (Maulana (Climate-Data.org). There are three reasons why & Setiawan 2018). choosing a small place like Sidoarjo Regency as the In the previous research, Setiawan and Herdi- research location, namely 1) It is easier to gener- ate basic data for the algorithm; 2) The large image anto (2018) created a mobile application that could capacity of the smartphone camera becomes ineffi- analyse and recommend the need for nitrogen in rice plants based on the rice leaves’ colour. In this appli- cient if applied to a large area, and 3) Unique land- cation, a set of process stages for image processing forms. The Jabon Subdistrict area is formed from and classification is implanted. It is used to analyse the river and sea sedimentation or fluvio-marine. 62 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 − 75 One of the characteristics of this landform is high had 4GB RAM and a 4,000 mAh battery. The image salinity because the material contains salt deposits. had a size of 4,000 × 3,000 pixels. Another tool used Many people take advantage of it by cultivating rice, was a 5 m long pole as a vertical photo-taking tool. although the harvest in recent years has decreased. Then there was the Gimbal for the terrestrial camera The same landform character is found in Central stabilizer (Astika et al. 2011). Java, Rembang Regency. It is also formed from the Field observation sedimentation of rivers and seas. Erosion material is The research location was determined based on deposited by rivers on the coast and combined with the initial salinity analysis with an Electrical Con- material carried by ocean waves (Wulan et al. 2016). ductivity value of 9.6 mS/cm and exchangeable Jabon district is located in a coastal lowland Na of 1.8 cmol/kg. This study was located in two area, which topography condition is influenced by transects. The length of Transect 1 was ± 4 km, and fluvio-marine sediment and alluvium material (Mar - Transect 2 was ± 3.4 km. Transect 1 was 316 ‒ 817 soedi et al. 1997). The soil types in the Jabon district meters, and Transect 2 was 104 ‒ 718 meters, where include Typic Hydraquents (Soil Survey Staff 2014). each transect had ten observation points. The salinity The rice field is an area of 1,883.86 ha, or 23.05% source’s distance was 10.65 km – point determina- (from the total area). Jabon district consists of two al- tion based on a free grid (Rayes 2007). Transect 1 luvial and marine landscapes. The alluvial landform consisted of observation points 1a to 10a, which with the alluvial plain sub-landform is in the western had a distance of 11.2 to 13.4 km from the salinity part of 2,417.01 ha or 29.6%. Marine landform with source. Additionally, transect 2 consisted of obser- plains tidal sub-land is east of Java Island, covering vation points 1b to 10b, which had a 10.3 to 13.5 km 5,756.34 ha or 70.4% (Marsoedi et al. 1997). distance from the salinity source. Material The point determination could be seen from soil A smartphone camera took the image in RGB analysis results (saline indication) and the coast dis- format with 48 megapixels. This terrestrial camera tance. The rice and soil samples were taken using an Figure 1. Distribution of research observation points 63 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61−75 active field survey method by constructing transects green, and blue). A gimbal smartphone camera sup- through the ArcGIS 10.3 application. Determina- ported the camera to stabilize, and the images were tion of the sampling point is taken based on field taken from a height of 5 m. The photoshoot was car- conditions / without a rigid grid and directly tagged ried out in the rice field with an area, and the land to take the location’s coordinates. Determination of adapted to the existing land with varying levels of observation points is done by purposive randomized leaf colour (Astika et al. 2011). The gimbal stabi- sampling (Rayes 2007). The soil and rice sampling lizer can stabilize the movement and disturbance of methods are described in "Soil and crops sampling" the wind because the actuator system in the gimbal section (page 65). The research location was deter- design uses a servo motor (Fahmizal et al. 2018). mined based on the age of the crops (in the vege- Servo motor is an electrical device used in machines tative phase). The age of rice crops obtained in the that function to push or rotate objects with high pre- survey activity was between the ages of 41 to 56 cision control in angle, acceleration, and speed. The DAP (days after planting). The plants were classi- angle of elevation of the gimbal will be controlled fied into the vegetative phase of class 2, rice crops stably (Suryana 2018). aged 41 – 64 DAP from the age data. Detection of The first step in taking photos using a smart- rice crops in rice fields with a good image is in phone, apart from making sure the camera is func- the vegetative phase 2 (LAPAN 2015). This stage tioning correctly, is paying attention to suitable also involved the preparation of tools and materi- weather conditions. Overcast skies or the hot sun als. Then the tools needed for the field survey were will affect the image. Smartphone photos require prepared, such as a trowel, SPAD (chlorophyll me- side lap, overlap, and image height settings so that ter), dreadlocks, 5 meter long sticks, administrative errors that occur due to movements such as tilt and maps of Jabon Regency, and other supporting maps poor lighting can be avoided (Syauqani et al. 2017). (Rayes 2007). The difference in the tilt and angle of sunlight, shooting is carried out simultaneously at 10:00 a.m. Photo-taking using smartphone camera and a minimum height of 5 meters for smartphone The tool used was a smartphone camera with camera use. Research conducted by Simanungkalit 48 megapixels and visible light wavelengths (red, 7.9 6.0 y = -0.0815x + 6.676 7.8 R² = 0.4964 5.9 7.7 5.8 7.6 5.7 7.5 y = -0.1355x + 9.2686 5.6 7.4 R² = 0.7299 7.3 5.5 10 11 12 13 14 10 11 12 13 14 Distance [km] Distance [km] 11.5 0.6 10.5 0.5 9.5 0.4 8.5 0.3 y = -0.0862x + 1.4586 y = -1.0276x + 21.852 7.5 R² = 0.6698 R² = 0.5376 6.5 0.2 10 11 12 13 14 10 11 12 13 14 Distance [%] Distance [km] Figure 2. Graph of soil salinity indicator. EC –Electrical Conductivity; ESP – Exchangeable Sodium Percentage; SAR – So- dium Adsorption Ratio; pH H O based on the distance from the salinity source ESP [%] EC [mS/cm] SAR pH H O 2 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 − 75 et al. (2019) found that smartphone photos obtained of nitrogen (Bremner 1996), phosphorus (Bray & Kurtz 1945) and potassium (Zakiyah et al. 2018) as from images taken at 10:00 a.m. had an aerial photo macronutrients. accuracy rate of above 95% in the perfect category. Salinity parameter analysis Soil and crops sampling In identifying macronutrients in rice crops in sa- The survey activities were carried out by taking line soils, salinity parameters must be considered. soil and crop samples. The soil and crop (leaf) sam- Salinity analysis was evaluated using the percentage ples were taken before or after aerial photographs of sodium exchange (ESP), soil pH, electrical con- on the same day. This step was taken so that the soil ductivity (EC), and sodium adsorption ratio (SAR) samples taken did not undergo significant changes (Djuwansah 2013). This parameter is called the sa- in the field conditions. Real-time sampling was linity indicator. Salinity parameter criteria are ex- also carried out so that the resulting data can have changeable sodium percentage (ESP) < 15% (Gupta high accuracy values. The soil sampling taken was & Sharma 1990), soil pH < 8.5 (Amran et al. 2015), a layer of the rice root area. The soil sampling was sodium adsorption ratio (SAR) < 13 (Robbins 1984), done on topsoil using the undisturbed soil sampling and electrical conductivity (EC) is 2 – 4 or > 4 mS/cm method (Putra & Nita 2020). The soil sampling was (Rhoades et al. 1989) in soil. Saline soil is different taken from a depth of 0 – 0.2 m (± 1 kg) (Vadas et from saline-sodic and sodic soil. Saline-sodic has al. 2006). The sampling of rice biomass was car- criteria of EC > 4 mS/cm, ESP > 15%, and pH > 8.5. ried out by taking part in the leaves. The determina- Sodic soil has criteria of EC < 4 mS/cm, ESP > 15%, tion of soil and crop sampling points based on field and pH > 8.5 (Sipayung 2003). conditions / without rigid grids and directly tagged to retrieve location coordinates using GPS. The de- Image pre-processing and digital number extraction cision of the observation points was carried out by The pre-processing of smartphone aerial photos purposive randomized sampling (Rayes 2007). The was an initial data processing process. There were composite soil sampling and rice leaf samples were several pre-processing stages performed (Muñoz conducted for laboratory analysis to obtain levels & Kravchenko 2011). The first stage was rectifica - 1.45 0.5 y = 0.0397x + 0.8255 1.40 R² = 0.609 0.5 1.35 1.30 y = 0.0554x - 0.19 0.4 R² = 0.5036 1.25 1.20 0.3 10 11 12 13 14 10 11 12 13 14 Distance [km] Distance [km] 5.0 4.5 4.0 3.5 3.0 y = 0.5397x - 2.5355 2.5 R² = 0.5036 2.0 10 11 12 13 14 Distance [km] Figure 3. Plant biomass analysis result and the distribution based on distance from saline source Note: N – nitrogen; P – phosphorus; K – kalium N [%] K [%] P [%] Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 −75 tion. The second step is to extract digital numbers in rected, and the digital number extraction was started ArcGIS 10.3 using the Extract Value to Points Tool by adding an improved aerial photo. (Putra & Nita 2020). This tool is used to obtain the Moreover, the digital number extraction was digital number of each pixel at the raster observation carried out. The extraction of digital numbers on point (Dell 2009). Digital numbers are RGB values a smartphone camera was carried out using each because rasters are made up of red, green, and blue rectified photo, and then a sampling point was en- tered. This study was conducted using visible (red, waves. In this study, the digital number value is the green, blue) channels obtained through shooting us- total value of the red, green, and blue waves in one ing smartphone cameras. pixel (RGB combination). The digital number trans- formation results were continued with statistical Statistical analysis analysis of correlation and regression using R soft- Initially, the laboratory data results and the ware and then compared with laboratory data cor- normality test result for potato production were related with rice crops. analyzed using R studio by the Shapiro-Wilk method The point of taking the digital number reclaimed (Royston 1992). The correlation test was used to aerial photo made diagonally. In one aerial photo, determine the closeness of the relationship between there were five digital number value retrieval points. variables and the direction of the relationship (Putra Each DN value retrieval point was three replications & Nita 2020) The correlation coefficient value (r) (15 points in total), then averaged into one aerial was compared to the r-table (Bewick et al. 2003). photo’s value (Astika et al. 2011). The smartphone camera aerial photos were entered into the ArcGIS Formulation and interpolation of macronutrients 10.3 application. The rectification was carried out deficiency in saline soil with the coordinates of the rectification coordinates The equation was used to estimate macronutrients in each aerial photograph corner. It was then cor- (nitrogen, phosphorus, and potassium) of rice crops in salinity stress. Moreover, the resulting equations were associated using an aerial photo with a raster 43 calculator on ArcMap 10.3 in ArcToolbox on the Map Algebra tool (Lubis 2011). The resulting interpolation estimated macronutrients (nitrogen, phosphorus, and potassium) in rice biomass. Accuracy assessment The validation test used a paired t-test to verify the correctness or certainty of a model. The reliabil- 10 11 12 13 14 ity test using R studio aimed to determine whether Distance [km] model consistency accuracy – validation using the Figure 4. Graph of chlorophyll index T-pair test (Montolalu & Langi 2018). RESULTS Soil salinity analysis results Based on Figure 2, the values of electrical con- ductivity (EC), SAR, ESP, and pH H O are getting bigger and closer to the source of salinity. The image y = 11.381x + 1.6929 R² = 0.7287 is included in the classification of moderate salinity values. The electrical conductivity obtained in this 10 11 12 13 14 study ranged from 7.35 to 7.92 mS/cm. According Distance [km] to Sipayung (2003) classification based on electri - Figure 5. Digital number distribution based on the distance cal conductivity values, moderate salinity is rated of the saline source Chlorophyll index Digital number Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 −75 at 4 – 8 mS/cm, where many plants are affected. It status of nutrients in leaves. Figure 3 has the lowest includes control in SAR because the value is < 13, chlorophyll value of 30.8 units/SPAD, and the high- ESP includes control because all points have ESP est is 41.38 units/SPAD. According to Prabowo et al. values < 15%, and pH < 8.5. That value corresponds (2018), chlorophyll measurement results with SPAD to the saline soil. values can be categorized into three criteria, name- The chemical properties used as criteria for sa- ly low (< 50), medium (50 ‒ 53), and height (> 53). line soils are characterized by soil EC more than The chlorophyll in this study had a SPAD value of 4/ > 4 mS/cm, EC (Electrical Conductivity) < 15%, < 50 came in the low category. The difference in rice and pH < 8.5 (Pardo 2010). The salinity indicator chlorophyll in various varieties was due to crops’ consisting of EC, pH H O, ESP, and SAR showed 2 ability to adapt to different salinity conditions. an increase in the value of each indicator as the dis- As shown in the graph in Figure 4, the chloro- tance from the saline source got closer (Figure 2). phyll’s value does not increase or decrease signifi- The EC values at all observation points increased to cantly. The salinity source’s far and proximity do 7.4 mS/cm at a distance of 13.5 km, then increased not affect the increase or decrease in chlorophyll’s to about 7.9 mS/cm at a distance of 10.5 km. The value. This condition is due to the difference in the same pattern of increase also occurred in pH H O age of crops and rice varieties used. and SAR. Soil H O pH increases from about 5.5 The average age of crops at a distance of 13.4 km to almost 6 as the salt sources are closer to each to 11.2 km from salinity source was 50 DAP (the day other. Another indicator is that the SAR increased after planting), with the youngest crops age 42 DAP. from 0.2% to 0.6%, and the ESP increased from 6 to Besides, in on-point observation, with a distance around 10. Therefore, the closer to the saline source, of 13.5 km to 10.3 km from the salinity source, the the higher the salinity level. This condition causes average age of crops was 54 DAP with the oldest the area near the source of salinity to become an area age of 58 days after planting and the youngest of of high salinity flow to crops. 50 DAP. The varieties used at all observation points Macronutrient biomass analysis results were Ciherang, Inpari 32, and Inpari 42 varieties. The chlorophyll index was reviewed from varie- Analysis of nitrogen, phosphorus, and potassium ties. The average value of chlorophyll in the Ci- biomass: herang variety was 36.1 units, Inpari 32 variety was The laboratory analysis of nitrogen, phosphorus, 37.6 units, and Inpari 42 was 38 units. According and potassium crops (leaves) showed that the total to (Banyo et al. 2013), in terms of crop life, crops nitrogen, phosphorus, and potassium levels of rice with a more extended planting period cause a higher crops decreased towards the saline source at the ob- chlorophyll concentration than plants that grow fast- servation point. The results of the study of the total er in the vegetative phase. According to (Muyassir nitrogen, phosphorus, and potassium values of rice 2012), crop age affects the value of chlorophyll in crops showed a decrease as the distance began to get leaves. Mardiansyah et al. (2018) stated that the Ci- closer to the source of salinity (Figure 3). Salinity is herang variety has moderate tolerant characters to very influential, where the higher the salinity value causes inhibition of nutrient absorption for crops. high salinity. Inpari 32 is an inbred variety from the Salinity interferes with crops’ nutrient uptake in two Ciherang selection, and Inpari 42 is a salinity toler- ways. First, the ionic strength of the substrate, re- ant variety (Suhartini & Zulchi 2018). Their results gardless of its composition, can influence nutrient stated that salinity did not affect the crop’s chloro- uptake and translocation. Second, salinity interferes phyll levels (Nurgayah & Irawati 2017). with plant mineral relationships by reducing nutrient Digital number extraction availability through competition with ions (Monica In this study, the digital number (DN) used is et al. 2014). a combination of RGB. Extraction of digital nu- Leaf chlorophyll index meric values is taken from the total values of the Chlorophyll is one of the factors to determine the red, green, and blue pixels visible in the aerial pho- 67 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 −75 to. Pixels (picture elements) are the minor element The highest digital number value is obtained at points in a photo. The numeric number (1 byte) of 162.8, while the lowest at 119.4. The highest digital a pixel is called the digital number (DN) (Efendi number value is obtained at 159.9, while the low- 2012). The use of smartphones can be implement- est at 121.8. The digital number’s value increases to ed independently (stand-alone) by storing data on point ten, the digital number value decreases. The the mobile device (for simple applications) (Gunita closer the salinity of the source, the digital number et al. 2013). The smartphone’s camera uses visible values decrease (Figure 5). RGB (red, blue, and green) electromagnetic waves. Statistical analysis result Unlike terrestrial cameras, UAVs, drones, and oth- ers, the DN value uses visible light (RGB), NIR or Normality test: SWIR, or Red Edge. However, if used a vegetation/ The observation variable carried out by the nor- soil index, the name is an index number, not a digi- mality test can be said to be normal if the p-value tal number (Bernardi et al. 2017). The use of smart- is p ≥ 0.05. The Digital Number smartphone cam- phones is developing satellite imagery and UAV era has a normality test value of 0.236. N total [%] research with a higher level of precision. Salinity crops have a normality test value of 0.340. Then, characteristics between locations are different, so it from the availability of nutrients in the biomass, is necessary to use smartphones (Astika et al. 2011). namely nitrogen total [%], crops have a normality The digital number extraction is started with add- test value of 0.340, phosphorus total [%] value of ing RGB photos. Then a digital number extraction 0.601, and potassium total [%] value of 0.604. From is performed. Digital number extraction on a smart- the availability of nutrients in the soil, namely ni- phone camera is done using each photo rectified and trogen total [%], the soil has a normality test value then inserted sampling point. In one aerial photo, of 0.246, P O total [mg/100g] value of 0.614 and 2 5 there are five digital number value retrieval points. K O total [mg/100g] value of 0.574. EC [mS/cm] Each DN value retrieval point consists of three re- has a normality test value of 0.627. Salinity indica- plays (15 points in total), then averaged to one aerial tors can be seen from pH H O data has a normality photo’s value. The data of digital number extraction test value of 0.734; pH KCl data has a normality results through aerial photos utilizing smartphone test value of 0.530, ESP value of 0.254, and SAR camera obtain results with details based on Figure 5. value of 0.298. All parameters data can be said to T a b l e 1 Correlation analysis of parameters with smartphone camera digital number DN Chlorophyll EC pH H O ESP SAR N [%] P [%] K [%] DN 1 –0.11 –0.81 –0.70 –0.78 –0.84 0.91 0.84 0.84 Chlorophyll 1 0.15 –0.12 0.10 0.06 –0.02 0.11 0.11 EC 1 0.63 0.83 0.92 –0.83 –0.69 –0.69 pH H O 1 0.55 0.60 –0.71 –0.84 –0.84 ESP 1 0.97 –0.87 –0.62 –0.62 SAR 1 –0.91 –0.71 –0.71 N [%] 1 0.87 0.87 P [%] 1 1.00 K [%] 1 Description: EC – Electrical Conductivity; ESP – Exchangeable Sodium Percentage; SAR – Sodium Adsorption Ratio; pH – acidity. Note: N – nitrogen; P – phosphorus; K – kalium 68 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 − 75 be expected because the value is more than 0.05. All The R value is obtained from the regression for- parameters can be said to be expected so that they mula in nitrogen, phosphorus, and potassium crops, can be continued to the correlation. which means the data is accurate. The regression equation in Figure 6 shows that the y-axis nitrogen, Correlation between salinity and biomass nutrient phosphorus, and potassium total [%] in rice crops, availability and the x-axis shows the smartphone camera dig- The correlation test between parameters and ital number (DN) value, so nitrogen, phosphorus, smartphone camera digital number values is pre- and potassium total in rice crops are affected by sented in Table 1. The macronutrients (nitrogen, the smartphone camera DN value. In Figure 6, val- phosphorus, and potassium) and salinity indicators ues 0.0035 (N), 0.0049 (P), and 0.0478 (K) are the in rice crops based on the smartphone camera digital slopes that determine linear regression direction and numbers are processed from smartphone regression 0.8192 (N), –0.2042 (P), and –2.6717 (K) is inter- test equations camera digital number values and the cept value. The slope value indicates a positive that results of the analysis of crops and soil samples in the higher the x-values than the greater the y-val- rice crops in the field. ue. The slope value also shows the rate of increase The equation used is the equations of smartphone of nitrogen, phosphorus, and potassium total rice camera digital numbers. Based on Table 1, the com- crops, an increase of nitrogen, phosphorus, and po- parison results show that smartphone camera digital tassium the total rice crops increased by 0.0035 (N), numbers calculate the r-value greater than the r-ta- 0.0049 (P), and 0.0478 (K). In contrast, the intercept ble (0.4438). It can be said that the DN smartphone value refers to the initial calculation value, when the camera value is increasing. The value of nitrogen, values x = 0, then nitrogen, phosphorus, and potassi- phosphorus, and potassium will also increase. The um total rice crops are 0.8192 (N), –0.2042 (P), and regression tests can be carried out and used to de- –2.6717 (K), respectively. termine the nutritional estimates of nitrogen, phos- The regression results have obtained the estima- phorus, and potassium in rice crops. In contrast, tion of nitrogen, phosphorus, and potassium nutri- chlorophyll has a low correlation value and negative ents in rice crops. Nitrogen, phosphorus, and potas- results on the DN smartphone camera. Chlorophyll sium data in rice crops in the field with estimated has a lower calculated r-value, so the correlation re- data nitrogen, phosphorus, and potassium using sult value cannot be performed for regression tests. smartphone camera DN show values were not much The macronutrients (nitrogen, phosphorus, and different. The DN smartphone camera on each aerial potassium) in rice crops based on salinity indicators photo after extraction produces pixels with red, blue, such as pH H O, EC (mS/cm), ESP, and SAR have and green values converted in a DN value that can processed the analysis of crops and soil samples in be used to guess the macronutrients of rice crops. rice crops in the field. Based on Table 1, the com- parison results show that the correlation between Accuracy assessment (t-pair test) nitrogen, phosphorus, and potassium rice crops on The estimation results using a smartphone were the salinity indicators such as pH H O, EC, ESP, and 2 tested using a paired T-test to see the similarity of SAR values is negative. It can be said that the high- the laboratory data with the estimated N, P, and K er the value of salinity indicators such as pH H O, 2 nutrients for rice crops and salinity indicators. The EC, ESP, and SAR will be inversely proportional estimated p-value is lower than p > 0.05, consist- to nitrogen, phosphorus, and potassium rice crops, ing of N, P, and K, which are 0.58, 0.81, and 0.97, namely decreasing. In comparison, chlorophyll has respectively. The t-value also shows no difference a low correlation value for controlling pH H O, ESP, 2 between the results of laboratory analysis and esti- EC, and SAR indicators. Chlorophyll has a lower mates, which are –0.45 (N), 0.19 (P), and 0.11 (K), calculated r-value, so the correlation result value respectively. The paired t-test shows that the calcu- cannot be performed for regression tests. lated t-value is smaller than the t-table value (0.68) Regression (R ) parameters nitrogen, phospho- for N, P, and K. This shows no difference between rus, and potassium total biomass rice using smart- the results of macronutrient analysis from the labo- phone camera digital number. 69 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 −75 ratory and the estimation results using a smartphone. value of the DN Smartphone camera will have a po- sitive effect on macronutrients (nitrogen, phospho- rus, and potassium) of 83.7% (nitrogen), 70.34% DISCUSSION (phosphorus), and 70.29% (potassium). In other words, the positive effect shows that the higher Smartphones can be used to identify macronu- the DN value of the Smartphone camera, the more trient biomass in rice with an estimated value that macronutrients (nitrogen, phosphorus, and potas- does not differ from laboratory analysis results. sium) increase. Jabon Regency has a fluvio-marine DN smartphone can be used to predict N, P, and K landform formed from a sedimentation process from crops. A study by (Amri & Sumiharto 2019) shows a mixture of river sediment (alluvium) and marine that a smartphone system can detect nitrogen, phos- sediment (marine). Jabon was previously an ocean phorus, and potassium nutrients in rice fields in the that has become land so that the area still contains Special Region of Yogyakarta. In addition, based on salt deposits in the underground part. LPT Bogor research, the results of the detection of The metabolic imbalance caused by ion (Na ) nutrient levels of nitrogen, phosphorus, and potassi- poisoning causes nutrient deficiency (nitrogen, um showed an average detection accuracy of 70.65% phosphorus, and potassium). The destructive effect (N 94.98%, P 50.84%, and K 66.14%). The best for- of salinity on plants is related to the high osmotic mula that results from the research results is NTotal pressure of water, an imbalance between Na and K, = 0.0035 * DN + 0.8192 (R 0.84), PTotal = 0.0049 Ca, Mg ions, and decreased uptake of nitrogen and * DN – 0.2042 (R 0.70) and KTotal = 0.0478 * DN phosphorus (Grattan & Grieve 1998). Salinity seems – 2.6717 (R 0.70), respectively (Figure 6). to affect two processes, namely water relations, and Salinity affects the concentration of macro- ionic relations. During initial exposure to salinity, nutrients in crops, reduces the accumulation of the crops experience water pressure, which reduces nitrogen in crops, phosphorus concentration, and the development of leaves. During long-term salinity decreases the accumulation of K in crop tissues. exposure, crops experience ionic stress, causing This equation shows that every 1 unit increase in the three potential effects on crops: reducing water po- 1.45 0.54 1.40 y = 0.0035x + 0.8192 R² = 0.837 1.35 0.46 1.30 0.38 y = 0.0049x - 0.2042 1.25 R² = 0.7034 0.30 1.20 118 128 138 148 158 118 128 138 148 158 Digital number Digital number 4.4 3.6 2.8 y = 0.0478x - 2.6717 R² = 0.7029 2.0 118 128 138 148 158 Digital number Figure 6. Regression graphic of nitrogen, phosphorus, and potassium total crops with DN smartphone camera Note: N – nitrogen; P – phosphorus; K – kalium N [%] K [%] P [%] Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 − 75 tency, direct toxicity of any absorbed Na and Cl, and increases P or does not affect it. It is not surpris- disruption to the absorption of essential nutrients ing that differences between studies occur because (Flowers & Flowers 2005). Salinity causes severe P concentrations vary significantly in different ex- damage to many cellular and physiological process- periments, and other nutritional interactions may es, including photosynthesis, nutrient absorption, coincide. water absorption, root growth, and cell metabolism, The increased salt concentration leads to the ac- which leads to decreased results (Darwish et al. cumulation of toxic ions such as Cl and, in particu- 2009). Soil control affects the absorption of nitro- lar, Na in the cytosols. Several studies have shown gen, phosphorus, and potassium in crops. The spe- that the concentration of K in crop tissue decreases cific effects of soil control on crop metabolism, EC along with increased salinity of NaCl . The decrease + - + + in leaf aging, are associated with Na and Cl ion in K content in crops by Na is a competitive pro- + + accumulation and decreased K . Salinity associat- cess. Salinity decreases the accumulation of K on ed with excess NaCl affects crop growth and yield leaves (Manchanda & Garg 2008). The adverse ef- by suppressing water and mineral absorption and fect of salinity on crops is associated with high wa- normal metabolism (Al-Karaki 2000). According ter osmotic pressure, the imbalance between Na ions to Sipayung (2003), salinity inhibits the growth of with K, Ca, Mg. Moreover, it is also associated with roots, stems, and leaf area, as well as metabolic im- decreased absorption of N and P (Grattan & Grieve balances caused by ion poisoning (Na ) and nutrient 1998). In general, salinity reduces the accumulation deficiency (nitrogen, phosphorus, and potassium). of N in crops. This is because a decrease in nitrate The P concentration in agronomy crops in the field concentration mainly accompanies the increase in decreases with increased salinity. Salinity decreases absorption and accumulation of chloride (Garg et al. P concentration in crop tissues; elsewhere, salinity 1993). 1 2 Figure 7. Results of aerial photo estimate nitrogen (1), phosphorus (2), and potassium (3) total biomass rice crops from the smartphone camera 71 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 − 75 The diversity of soil salinity is affected by the in- The most significant salinity factor in Jabon is the fluence of the source of soil salinity, namely seawa- geographical position of Sidoarjo, which is on the ter in coastal areas. Thus, the observation point 1a to seafront. Salinity as the most significant factor causes 10a and 1b to 10b describe salinity variability from a decrease in nitrogen, phosphorus, and potassium low to high. Marwanto et al. (2009) explained that in rice crops in Jabon. Meanwhile, other factors the closer to the saline source (coastal), the salinity such as soil texture, plant varieties, and suboptimal increases. The digital signal processing is obtained management are less influential than the effect of sa- by making use of the characteristic wavelength of linity. Salinity affects soil texture, soil structure, and the reflected leaves. The smartphone camera sensors uptake of plant nutrients. With high salinity values, can measure, analyse, monitor a condition, and then the availability of plant nutrients is deficient, so that ‘r’ EC and changes in its surroundings. Several sen- the macronutrient of rice has decreased. sors on a smartphone camera include ambient light sensors, temperature, and humidity sensors (Mau- LIMITATIONS AND PROBLEMS IN THE FIELD lana & Setiawan 2018). The smartphone camera is Aerial photos taken using a long stick with an active detection example that provides its energy a smartphone camera are not used for large land due source to illuminate targets and uses sensors to mea- to the viewing angle. If the entire expanse of the rice sure reflection energy by calculating the reflection field in the photo with a camera using a long stick, angle or time required to recover energy. Some of then some land will be photographed from the side. the smartphone cameras utilize visible RGB (red, Aerial photography is affected by sunlight, brighter blue, and green) electromagnetic waves. Partially, cross-sections, reflections of light from rice fields, RGB does not affect salinity; the digital number and shadows in aerial photographs (such as a long (DN) used is a combination of RGB. The results stick). Taking photos needs to ensure good weath- of the smartphone photo camera refer to the RGB er conditions. Avoid taking pictures when the sky is image. The smartphone DN extraction value is ob- cloudy or the sun is scorching. tained from the total RGB value in pixels (RGB The pick-up is also affected by the area of the rice combination). A combination of RGB values (red, field map. As in transacts 2, observation points 1 and green, and blue) of aerial photos combine the three 2 have a small map width so that the outer part of primary colours, resulting in various colours in one the map is visible on the aerial photos. Choose a site (Santoso & Handoyo 2015). with a large rice field map. Because of the remote The development of technology in various fields location, it cannot be done for extrapolation to the impacts digital image processing, one of them is other place. However, it can be overcome by the in- on smartphones. Smartphones have many features, terpolation tool in ArcGIS software. One of them is such as digital image capture (Budiman et al. 2019). Raster Calculator Tool using a math operation. This A digital image is composed of a collection of dots tool will calculate each pixel using an expression/ called pixels to form a digital photo. Smartphones algorithm (Rogers & Staub 2013). The smartphone are operated by a Linux-based operating system that camera device used in this study cannot connect includes an operating system, middleware, and ap- with other smartphone cameras or remote control. It plications (Safaat 2011), more than 80% of smart- can be seen the accuracy of aerial photo-taking. So phone users. Smartphones include operating sys- it uses a timer, then it must be repeated in lifting the tems with open source licenses that everyone can long stick. Use a smartphone that supports remote use freely to support daily activities and work, in- control applications. cluding in agriculture (Setiawan & Herdianto 2018). The variety and age of rice crops in the field vary The use of smartphone cameras is one of the devel- considerably at each observation point, allowing opments of satellite and drone imagery. The advan- variations in rice plants’ appearance. The location tages are good image quality and can to be arranged is determined by trying to make the variety and and easy to carry everywhere and the results are age of the rice crop more uniform. The optimal age fast, there are menus such as brightness, sharpen, of rice crops used for shooting is in the vegetative smooth, and edge detection (Adiyat 2013). phase 2 because it produces a better image (Mosleh 72 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61 − 75 et al. 2015). In processing aerial photos using the Disclosure statement. The authors declare no smartphone camera, the extraction is still in com- conflict of interest. posite data. Because it is done manually, the height is not standardized. Even though the gimbal has REFERENCES been installed, the shock when taking also affects the shooting results. A sampling at four corners and Adiyat, I. (2013). Flash-based digital image processing appli- one center may be fewer DN and chlorophyll sam- cation on mobile devices. Pontianak: Universitas Tanjung- pura, 49p. ples of rice crops. Aerial photo processing using Ahmed, M.F. and M. Z. Haider. (2014). Impact of salinity on a smartphone camera extraction can be done by pro- rice production in the south-west region of Bangladesh. En- gramming to get a DN value that includes all pixels vironmental Science an Indian Journal, 9(4), 136 – 141. Al-Karaki, G.N. (2000). Growth, water use efficiency and or can do orthophoto, then apply a zonal function. In sodium and potassium acquisition by tomato cultivars taking aerial photos using a smartphone camera, the grown under salt stress. Journal Plant Nutrition, 23, 1 – 8. DOI:10.1080/01904160009381992. handling and height must be the same as the effec- Amran, M.B., Sari, N.K.E., Setyorini, D.A., Wahyu, Y., Wid- tive area and more stable in the shooting. iani, D. and Irnameria, D. (2015). Soil quality analysis of Sawarna Beach, Lebak Regency, Banten Province. Pro- ceedings of the National Symposium on Science Innovation and Learning, 8. CONCLUSIONS Amri, M.M. and Sumiharto, R. (2019). Nitrogen, phospho- rus, potassium measurement system with local binary pat- tern and regression analysis. IJEIS (Indonesian Journal The content of soil macronutrients (nitrogen, of Electronics and Instrumentation Systems), 9(2), 107. phosphorus, and potassium) significantly affects the DOI:10.22146/ijeis.34132. Astika, I.W., Suud, H.M., Setiawan, R. P.A., Syuaib, M.F. rice crop biomass. In saline soils, the availability of and Solahudin, M. (2011). Estimation of rice productivity these macronutrients is reduced, problematic and with image processing taken from mini airplanes. National Seminar PERTETA, 106 – 112, http://repository.ipb.ac.id/ hampered. Estimating macronutrient content using handle/123456789/68618, Mechanical & Biosystem Engi- a terrestrial camera is an effort to identify nutrient neering . deficiencies early. The macronutrient content has Bernardi, A.C. de C., Grego, C.R., Andrade, R.G., Rabello, L.M. and Inamasu, R.Y. (2017). Spatial variability of veg- a strong effect on the digital number of smartphone etation index and soil properties in an integrated crop-live- camera values. It has a positive regression (nitro- stock system. Revista Brasileira de Engenharia Agrícola e Ambiental, 21(8), 513 – 518. DOI:10.1590/1807-1929/agri- gen, phosphorus, and potassium) of 0.84, 0.70, and ambi.v21n8p513-518. 0.70. This result means that the higher the saline in- Banyo, Y.E., Nio, A.S., Siahaan, P. and Tangapo, A.M. (2013). dicator’s value, the lower the nitrogen, phosphorus, Rice leaf chlorophyll concentration during water deficien- cy induced with polyethylene glycol. Jurnal Ilmiah Sains, and potassium values in rice crops. A smartphone 13(1), 1 – 8. DOI:10.35799/jis.13.1.2013.1615. camera (terrestrial) can monitor salinity stress’s Bewick, V., Cheek, L. and Ball, J. (2003). Statistics review 7: Correlation and regression. Critical Care, 7(6), 451. impact on reducing macronutrients (nitrogen, phos- DOI:10.1186/cc2401. phorus, and potassium) in rice crops. The reliabil- Bray, R.H. and Kurtz, L. (1945). Determination of total, organ- ic, and available forms of phosphorus in soils. Soil Science, ity test results of an aerial photo’s digital number 59(1), 39 – 46. value through a smartphone camera have a smaller Bremner, J.M. (1996). Nitrogen-total. Methods of Soil Analysis: t-count value than the t-table. The estimated value Part 3 Chemical Methods, 5, 1085 – 1121. DOI:10.2136/ sssabookser5.3.c37. is not significantly different from the actual data, Budiman, A., Utomo, P. and Rahayu, S. (2019). Development so the aerial photo’s digital number value through of applications for early detection of rice pest attacks based on android. Journal of Applied Abdimas, 4(1), 33 – 39. the smartphone camera can be applied to guess the DOI:10.25273/jta.v4i1.3805. macronutrients nitrogen, phosphorus, and potassi- Central Statistics Agency (2020). Harvest area, production, and um rice crops and salinity indicators. rice productivity by province 2018 – 2019. Retrieved from the Central Bureau of Statistics: https://www.bps.go.id/sub- ject/53/tanamanpangan.html#subjekViewTab. Acknowledgements. The author would like to Climate-data.org. (2010). Climate data for cities worldwide. Retrieved from https://en.climate-data.org/. thank the Faculty of Agriculture Universitas Braw- Darwish, E., Terterink, C., Khalil, M., El-Shihy, O. and Munnik, ijaya who has supported research activities through T. (2009). Phospholipid signaling responses in salt-stressed the 2020 LPPM UB Beginner’s grant. rice leaves. Plant and Cell Physiology, 50(5), 986 – 997. DOI:10.1093%2Fpcp%2Fpcp051. 73 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61− 75 Dell, M. (2009). GIS analysis for applied economists. Depart- tions. ULTIMATICS: Journal of Informatics Engineering, ment of Economics, Massachusetts Institute of Technology. 10(2), 85 – 92. DOI:10.31937/ti.v10i2.957. Djuwansah, M. (2013). Sodium status in polluted soil by tex- Monica, N., Vidican, R., Rotar, I., Stoian, V., Pop, R. and Mi- tile industry waste in Rancaekek, Bandung Regency. clea, R. (2014). Plant nutrition affected by soil salinity and Journal of Soil and Climate, 37(1), 10. DOI:10.21082/jti. response of rhizobium regarding the nutrients accumula- v37n1.2013.25-34. tion. BioFlux; ProEnvironment, 7, 71 – 75. Efendi, R. (2020). Design and build augmented reality applica- Montolalu, C. and Langi, Y. (2018). The effect of basic com- tions for android-based detection of medicinal plants. Jour- puter and information technology training for teachers with nal of IKRA-ITH Informatics, 4(1), 11. paired t-test (Paired Sample T-Test). D’CARTESIAN, 7(1), Fahmizal, F., Dewantama, G.Y., Pratama, D.B., Fathuddin, F., 44 – 46. DOI:10.35799/dc.7.1.2018.20113. and Winarsih, W. (2018). Design a camera stabilizer system Motamed, M., Asadi, K.R. and Amiri, E. (2008). Response of (gimbal) with fuzzy logic for photo and video capture. Jour- high yielding rice varieties to NaCl salinity in greenhouse nal of Information Technology and Computer Science, 5(3), circumstances. African Journal of Biotechnology, 7(21), 277 – 286. DOI:10.25126/jtiik.201853785. 3866 – 3873. Flowers, T.J. and Flowers, S.A. (2005). Why does salinity Mosleh, M.K., Hassan, Q.K. and Chowdhury, E.H. (2015). Ap- pose such a difficult problem for plant breeders? Agricul- plication of remote sensors in mapping rice area and fore- tural Water Management, 78, 15 – 24. DOI:10.1016/j.ag- casting its production: A review. Sensors, 15(1), 769 – 791, wat.2005.04.015. DOI:10.3390/s150100769. Garg, B., Vyas, S., Kathju, S., Lahiri, A., Mali, P. and Shar- Muñoz, J.D. and Kravchenko, A. (2011). Soil carbon mapping ma, P. (1993). Salinity-fertility interaction on frowth, using on-the-go near infrared spectrossalinity, topogra- mineral composition and nitrogen metabolism of Indian phy and aerial photographs. Geoderma, 166(1), 102 – 110. mustard. Journal of Plant Nutriotion, 16, 1637 – 1650. DOI:10.1016/j.geoderma.2011.07.017. DOI:10.1080/01904169309364639. Muyassir, M. (2012). Spacing effect of plantation, age and num- Grattan, S.R. and Grieve, C.M. (1992). Mineral element acqui- ber of seeds on yield of rice (Oryza sativa L.). Journal of sition and growth response of crops grown in saline envi- Land Resource Management, 1(2), 207 – 212. ronments. Agriculture, Ecosystems & Environment, 38(4), Nurgayah, W. and Irawati, N. (2017). The relationship between 275 – 300. DOI:10.1016/0167-8809(92)90151-Z. chlorophyll-A and physicochemical parameters in the wa- Grattan, S.R. and Grieve, C.M. (1998). Salinity–mineral nutri- ters of the village of Tanjung Tiram, North Moramo Dis- ent relations in horticultural crops. Scientia Horticulturae, trict, South Konawe Regency. Sapa Laut, 2(4), 97 – 102. 78(1 – 4), 127 – 157. DOI:10.1016/S0304-4238(98)00192- DOI:10.33772/jsl.v2i4.3817. 7. Pardo, J. (2010). Biotechnology of water and salinity stress tol- Gunita, M.H., Andri, S. and Bandi, S. (2013). Mobile GIS- erance. Current Opinion in Biotechnology, 21(2), 185 – 196. based digital map location marker application on android DOI:10.1016/j.copbio.2010.02.005. smartphone. Journal of Geodies, 2(4), 26 – 39. Prabowo, R.Y., Rahmadwati, R. and Mudjirahardjo, P. (2018). Gupta, S. and Sharma, S. (1990). Response of crops to high Classification of nitrogen content based on leaf color exchangeable sodium percentage. Irrigation Science, 11(3), through color clustering using fuzzy C means and hybrid 173 – 179. DOI:10.1007/BF00189455. PSO K-means methods. EECCIS Journal, 12(1), 1 – 7. Iswari, A.R., Hani’ah, H., dan Nugraha, A.L. (2016). Analysis Putra, A.N. and Nita, I. (2020). Reliability of using high-reso- of rice production fluctuations due to the effect of drought lution aerial photography (red, green and blue bands) for in Demak Regency. Journal of Geodesy Diponegoro Uni- detecting available soil water in agricultural land. Jour- versity, 5(4), 80404. nal of Degraded and Mining Lands Management, 7(3), Kementerian Pertanian (2019). Data for the last five years of 2221 – 2232. DOI:10.15243/jdmlm.2020.073.2221. rice production and harvest area. Jakarta: Ministry of Agri- Putri, R.E., Yahya, A., Adam, N.M. and Aziz, S.A. (2016). Vari- culture Republic of Indonesia. ability of rice yield with respect to crop health. Jurnal Te- LAPAN. (2015). Guidelines for monitoring rice plant growth knologi, 78, 1–2. DOI:10.11113/jt.v78.7272. phase using remote sensing satellite data. LAPAN, Jakar- Rachman, A., Dariah, A. and Sutono, S. (2018). Management ta: Center for Remote Sensing Utilization of the National of high-salt saline rice fields. Jakarta: IAARD Press, pp. Aeronautics and Space Institute, pp. 2 – 5. 20 – 48. Lubis, R.S. 2011. Estimation of correlation between soil char- Rayes, M.L. (2007). Land Resource Inventory Method. Yogya- acteristics and carbon stock in secondary forest. Tropical karta: ANDI Yogyakarta. Silviculture Journal, 3(1), 78. Rhoades, J., Manteghi, N., Shouse, P. and Alves, W. (1989). Manchanda, G. and Garg, N. (2008). Salinity and its effects on Soil electrical conductivity and soil salinity: New for- the functional biology of legumes. Acta Physiologiae Plan- mulations and calibrations. Soil Science Society of tarum, 30(5), 595 – 618. DOI:10.1007/s11738-008-0173-3. America Journal, 53(2), 433 – 439. DOI:10.2136/ss- Mardiansyah, M., Palupi, T. and Maulidi, M. (2018). Response saj1989.03615995005300020020x. of several local rice varieties to salinity stress in the nursery Robbins, C.W. (1984). Sodium adsorption ratio-exchange- phase. Agricultural Student Science Journal, 7(3), 1 – 9. able sodium percentage relationships in a high potassi- Marsoedi, D., Pusat P.T., Widagdo, A., Dai, J., Suharta, N., um saline-sodic soil. Irrigation Science, 5(3), 173 – 179. Darul, S.W.P., Hardjowigeno, S., Hof, J., & Jordens, E.R. DOI:10.1007/BF00264606. (1997). Guidelines for landform classification. Puslittanak, Rogers, S.R. and Staub, B. (2013). Standard use of geograph- Bogor: Center for Soil and Agroclimate Research, 33p. ic information system (GIS) techniques in honey bee re- Marwanto, S., Rachman, A., Erfandi, D. and Subiksa, I.G. search. Journal of Apicultural Research, 52(4), 1 – 48. (2009). Soil salinity level in intensive paddy fields in Indra- DOI:10.3896/IBRA.1.52.4.08. mayu Regency, West Java. Bogor: Soil Research Institute, Royston, P. (1992). Approximating the shapiro-wilk W-test for pp. 28 – 34. non-normality. Statistics and Computing, 2(3), 117 – 119. Maulana, S.H. and Setiawan, E.B. (2019). Utilization of sensors DOI:10.1007/BF01891203. on android smartphones for plant nursery recommenda- Ruminta, R. (2016). Analysis of decreasing production of rice 74 Agriculture (Poľnohospodárstvo), 67, 2021 (2): 61− 75 due to climate change in Bandung district West Java. Kulti- nal, 2, 1196p. vasi, 15(1), 37 – 45. Suryana, H. (2018). Design 3-axis gimbal as camera stabilizer Safaat, H. (2011). Androids; android based smartphone and while driving based on Proportional, Integral, Derivative tablet PC mobile application programming. Bandung: In- (PID) control system. Bandung: UIN Sunan Gunung Djati. formatics, 49p. 113p. Santoso, S. and Handoyo, E.D. (2015). The combination of us- Syauqani, A., Subiyanto, S. and Suprayogi, A. (2017). The ef- ing the Color model in detecting the location of the lips in fect of variations in flight height using the unmanned aerial color digital images. Journal of Informatics and Information vehicle (UAV) quadcopter DJI PHANTOM 3 pro on or- Systems Engineering, 1(2). DOI:10.28932/jutisi.v1i2.574. thophoto map making (case study at Diponegoro Univer- Sayed, H.E. and Sayed, A.E. (2013). Exogenous application of sity). Journal of Geodesy at Diponegoro University, 6(1), ascorbic acid for improve germination, growth, water rela- 249 – 257. tions, organic, and inorganic components in tomato (Lycop- USDA, A. (2020). United States Department of Agriculture ersicum esculentum Mill.) plant under salt-stress. New York (USDA). Retrieved from https://www.usda.gov/topics/ Science Journal, 6(10), 123 – 139. data/. Sembiring, H. and Gani, A. (2010). Adaptation of rice varieties Utama, M.H. (2015). Rice cultivation on marginal land: tips to on tsunami affected soils. Jakarta: Center for Rice Research. increase rice production. Yogyakarta: CV. ANDI OFFSET, Retrieved from www.dpi.nsw.gov.au/. pp. 105 – 208. Setiawan, E.B. and Herdianto, R. (2018). The use of android Vadas, P., Mallarino, A. and McFarland, A. (2006). The im- smartphones as a tool for analysis of nitrogen requirements portance of sampling depth when testing soils for their po- in rice plants. National Journal of Electrical Engineering tential to supply phosphorus to surface runoff. Extension and Information Technology (JNTETI), 7(3), 273 – 280. Fact Sheets, 1. DOI:10.22146/jnteti.v7i3.435. Widada, S. (2007). Symptoms of sea water intrusion in Simanungkalit, K., Damanik, M.R.S. and Lubis, D.P. (2019). the coastal area of Pekalongan City. Protobiont, 12(8), Optimization of unmanned aerial vehicle (UAV) aerial pho- 101 – 105. tography as remote sensing learning media. Tunas Geografi, WorldAgriculturalProduction.com. (2020). Production rice 8(1), 45 – 58. DOI:10.24114/tgeo.v8i1.15507. world 2020. Retrieved from http://www.worldagricultur- Sipayung, R. (2003). Salt stress and plant tolerance mecha- alproduction.com/crops/rice.aspx. nisms. Medan: University of Northern Sumatra, pp.1 – 6. Zakiyah, Z.N., Rahmawati, C. and Fatimah, I. (2018). Analy- Soussi, M., Ocana, A. and Lluch, C. (1998). Effects of salt stress sis of phosphorus and potassium levels in organic fertilizer on growth, photosynthesis and nitrogen fixation in chick- in the integrated Laboratory of Jombang District Agricul- pea (Cicer arientinum L.). Journal of Experimental Botany, ture Office. Indonesian Journal Of Chemical Research, 49(325), 1329 – 1337. DOI:10.1093/jxb/49.325.1329. 3(2), 38 – 48. DOI:10.20885/ijcr.vol3.iss2.art1. Soil Survey Staff. (2014). Key soil taxonomy, twelfth edition. Received: February 2, 2021 USA: United States Department of Agriculture, Natural Accepted: June 2, 2021 Resources Conservation Service, pp. 135 – 138. Suhartini, T. and Zulchi, T. (2018). Tolerance of Local Rice Germplasm to Salinity. Agroecotechnology Online Jour-
Agriculture – de Gruyter
Published: Jul 1, 2021
Keywords: remote sensing; visible light; android; nitrogen; phosphorus; potassium; salinity
Access the full text.
Sign up today, get DeepDyve free for 14 days.