Republic of the Philippines | Department of Education

Region III - Central Luzon | Schools Division of Bulacan

Assemblywoman Felicita G. Bernardino Memorial Trade School
Lias, Marilao, Bulacan

Senior High School Department

CORNELL NOTES 1

Class Number: Name: Capicio, Jhanina Krizzah B. Score:

Strand & Section: 11-STEM 2 Date: Parent’s Signature:

TITLE: Determination of Soil Nutrients and pH level using Image Processing and Artificial Neural Network.
KEYWORDS NOTES

INTRODUCTION According to Puno, J. C., Sybingco, E., Dadios, E., Valenzuela, I., & Cuello, J. (2017), the Philippines is
one of the Southeast Asian nations with fertile soil, leading to nutritional deficits in the soil which affects
the yield of crops. To ensure optimal yield, the soil must be prepared to ensure the crop receives the
proper nutrients.

Research Problem 1. The continued crop planting causes the soil to start losing nutrients, which also reduces the
quantity of crop that can be produced (Puno, J. C., Sybingco, E., Dadios, E., Valenzuela, I., &
Cuello, J. 2017).
2. The use of fertilizer, whether excessively or insufficiently, has a major impact on the production
of the agricultural sector (Nardozzi, C. 2017).
3. Adding inorganic fertilizer to the soil to increase production might have negative effects as these
chemicals affect the soil's nutritional content gradually (Rajasekharan, P., Nair, K.M., Rajasree,
G., & Kutty, M. C. N. 2013).

Soil Test Kit According to Puno, J. C., Sybingco, E., Dadios, E., Valenzuela, I., & Cuello, J. (2017), a quick way to
determine the soil's macronutrient content is to use a soil test kit.

Research Aims The study aims to develop a Soil Nutrient Analyzer using color recognition and the Rapid Soil Testing and
Soil Test Kit from the Bureau of Soils and Water Management (Puno, J. C., Sybingco, E., Dadios, E.,
Valenzuela, I., & Cuello, J. 2017). According also to the researchers, the software will use image
processing and an artificial neural network running on MATLAB to identify six different nutrient levels -
Nitrogen, Phosphorus, Potassium, Zinc, Calcium, and Magnesium - and the pH of the soil.

Using the Bureau of Soils and Water Management's Soil Test Kit and Rapid Soil Testing, image capture
Methodology from the controlled-light module box, image processing and artificial neural network using MATLAB, and
the outcome of the soil testing, which is the level of nutrients and pH of the soil, are all shown in the
overall framework (Puno, J. C., Sybingco, E., Dadios, E., Valenzuela, I., & Cuello, J. 2017).

Research Design Quantitative

General Framework in 1. Soil Testing - To determine the nutrients and pH level of the soil, Soil Test Kits (STK) and
accordance with the study of Rapid Soil Testing (RST) requires separate sets of chemicals. The chemicals and soil sample
Puno, J. C., Sybingco, E., will be combined according to the kit's instructions.
Dadios, E., Valenzuela, I., & 2. Image Capturing - Three photographs will be required to get the final soil testing result. One for
Cuello, J. (2017). the initial pH, one for the minerals zinc, calcium, and magnesium, as well as one for the
element’s nitrogen, phosphorus, and potassium. There will be a particular place to set the test
tube containing the soil and chemical combination for each parameter that will be checked
before image capture.
3. Image Processing - This section is divided into four parts: image capture, image
preprocessing, image segmentation, and feature extraction. The camera in the controlled-light
module box was used to take pictures.
4. Neural Network -The most important step is to improve the accuracy of the program's output
by using backpropagation neural network architecture to capture non-linear interactions
between parameters in a complex system. The input will be previously edited photographs and
the pH and nutrient level evaluations will be based on the bureau of soils and water
management standards (Cheriet, M., Said, J. N., and Suuen, C. Y. 1998), (Padmasree, P., &
Maheswari, R. 2013) and (Goh, A. T. C. 1995).
5. Result - The results will show the pH, relative amounts of nitrogen, phosphorus, potassium,
zinc, calcium, and magnesium in the soil.
Design Considerations As noted by Puno, J. C., Sybingco, E., Dadios, E., Valenzuela, I., & Cuello, J. (2017), first, the approach
recommended by the bureau of soils and water management's soil test kit and quick soil testing processes
will be carried out. The light-controlled module box will take pictures of how the chemical changes color
after reacting with the soil. MATLAB software will next be used to process the obtained image. It will
determine the soil sample's nutrients and pH level based on the outcome (Puno, J. C., Sybingco, E.,
Dadios, E., Valenzuela, I., & Cuello, J. 2017).

MATLAB An integrated development environment (IDE) for image processing is provided by MATLAB. Picture
capture, picture segmentation, feature extraction, and feature comparison are the four steps in the
procedure (Puno, J. C., Sybingco, E., Dadios, E., Valenzuela, I., & Cuello, J. 2017).

Data Gathered The researchers gathered number of images or pixel gathered per nutrient pH level.

Samples - The researchers used images of soil that had undergone soil testing to determine its pH and
nutrient levels. These included images of the nutrients nitrogen (100 images equivalent to low,
15 images equivalent to medium, and 15 images high), phosphorus (2000 pixels for low, 3500
pixels for medium, and 3500 pixels for high), potassium (128 equivalent to sufficient and 2 for
deficient), zinc (20 equivalent to sufficient and 30 for deficient), calcium (15 for sufficient and 35
for deficient), and magnesium (18 images equivalent to sufficient and 32 for deficient) of soil
after it undergo soil testing from the laboratory of BWSM.
- The evaluation of the software employed ten (10) randomly chosen soil samples with various
nutrients.

Summary of the Results Based on the investigtions of Puno, J. C., Sybingco, E., Dadios, E., Valenzuela, I., & Cuello, J. (2017), the
nutrients, pH level, and accuracy of six different nutrients were all 10 out of 10 (100% accuracy).

Conclusions MATLAB was used to effectively develop the software that analyzes images to identify the six nutrients
(nitrogen, phosphorus, potassium, zinc, calcium, and magnesium) and pH level of the soil. The research's
outcomes and conclusions led to the finding of the project was accurate and completely functional. The
project was finished and carried out effectively.

Recommendations of the 1. Gather more soil samples for the data sets so that they can produce more precise findings.
researchers 2. Have a recommendation for fertilizer depending on the results of the soil testing for a particular
crop.

Authors 1. John Carlo Velasco Puno

2. Elmer P. Dadios
3. Edwin Sybingco
4. Ira Valenzuela

Published Date August 13, 2018

References:

Puno, J. C., Sybingco, E., Dadios, E., Valenzuela, I., & Cuello, J. (2017). Determination of soil nutrients
and pH level using image processing and artificial neural network. In 2017IEEE 9th International
Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control,
Environment and Management (HNICEM) (pp. 1-6). IEEE.

Nardozzi, C. (2017). How to Test Your Soil - dummies.

http://www.dummies.com/home-garden/gardening/how- to-test-your-soil/.

Rajasekharan, P., Nair, K.M., Rajasree, G., & Kutty, M. C. N. (2013). Soil Fertility Assessment and
Information Management for Enhancing Crop Productivity in Kerala. Kerala State Planning Board.

Cheriet, M., Said, J. N., and Suuen, C. Y. (1998). A recursive thresholding technique for image
segmentation. Image Processing IEEE Transactions on, vol. 7, no. 6, pp. 918-921.

Padmasree, P., & Maheswari, R. (2013). A Novel Technique using Back Propagation in Neural Network.
International Journal of Computer Science and Engineering Technology (IJCSET).

Goh, A. T. C. (1995). Back-propagation neural networks for modeling complex systems. Artificial
Intelligence in Engineering, vol. 9, no. 1, pp. 143-151.
Republic of the Philippines | Department of Education
Region III - Central Luzon | Schools Division of Bulacan
Assemblywoman Felicita G. Bernardino Memorial Trade School
Lias, Marilao, Bulacan