A Computer Vision Based System to Monitor Crop health

Project Title

A Computer Vision Based System to Monitor Crop health

Project Area of Specialization Computer ScienceProject Summary

Determination of crop health is very important in agriculture. The right amount of fertlizer will help in efficient production of crops.To reduce the risk of yield-limiting deficiencies or costly overfertilizing, an instrument "SPAD" is avaialble which detects right amount of chrolorophyll [1]. But in a developing country like Pakistan most of our farmers are poor and are facing adverse poverty. They cannot afford such costly instruments. Our research based project is on detecting amount of chlorophyll present in a crop by examining its leaf using computer vision based techniques and algorithms. The amount of chlorophyll will tell about the required amount of fertilizer that the crop needs for better health and production. A method will be used in which principal component analysis is applied to digital images to calculate a greenness index using RGB components of the color image, which yields an estimate of the amount of N in the plant. The RGB features are extracted from the image and correlated with the SPAD values. To evaluate its quality, we will calculate the correlation between the index and measurements obtained with a SPAD502 chlorophyll meter, normally used in decision-making in fertilizer management. The RGB-based digital image analysis has the advantage over conventional subjective methods for being objective, fast, non-invasive, and inexpensive. The difference in chlorophyll value is used to calculate the required amount of fertilizer [2,3]. This will be a smart phone application and as most of the people have access to smart phones so it will surely make a differenece.

References:

1. Casadesús, Jaume, and Dolors Villegas. "Conventional digital cameras as a tool for assessing leaf area index and biomass for cereal breeding." Journal of integrative plant biology 56, no. 1 (2014): 7-14.

2. Pagola, Miguel, Rubén Ortiz, Ignacio Irigoyen, Humberto Bustince, Edurne Barrenechea, Pedro Aparicio-Tejo, Carmen Lamsfus, and Berta Lasa. "New method to assess barley nitrogen nutrition status based on image colour analysis: comparison with SPAD-502." Computers and electronics in agriculture 65, no. 2 (2009): 213-218.

3. Pérez-Patricio, Madaín, Jorge Luis Camas-Anzueto, Avisaí Sanchez-Alegría, Abiel Aguilar-González, Federico Gutiérrez-Miceli, Elías Escobar-Gómez, Yvon Voisin, Carlos Rios-Rojas, and Ruben Grajales-Coutiño. "Optical method for estimating the chlorophyll contents in plant leaves." Sensors 18, no. 2 (2018): 650.

Project Objectives

Chlorophyll is a pigment that gives plants their green color, and it helps plants create their own food through photosynthesis. Measuring the nitrogen nutrition status of plants is useful for nitrogen fertilizer management. As nitrogen is one of the main structural components of chlorophyll, its nutrition status is highly correlated with the greenness of leaves. A sustainable management of nitrogen fertilizer requires the knowledge of nitrogen nutrition status of crops. The SPAD 502 Plus Chlorophyll Meter instantly measures chlorophyll content or “greenness” of your plants to reduce the risk of yield-limiting deficiencies or costly overfertilizing. But SPAD is very expensive and most of our farmers cannot afford it. To accurately predict the amount of fertilizer required by a crop for proper functioning has always been a major concern for the farmers. Accurate prediction of nutrient requirements of plants during the cultivation period is necessary for efficient fertilizer use. The main objective of our project is to develop an efficient and cost-effective system so that it is approachable to our farmers. It will improve our agricultiural industry and avoid the risk of of yield-limiting deficiencies or costly overfertilizing.

Project Implementation Method

For the implementation purpose, we need smart phones with different resolution cameras. Photographs from the central zone of the youngest fully developed leaves are taken. Leaves are covered with a piece of fine suface black board from which a square of 1 cm2 is removed from the centre. Initially we will remove noise from image if there is any. Then the desired part of leaf is cropped from image by using image thresholding/filtering techniques. A total of eight different greenness indices are calculated resulting from the combination of two ways to analyse colour (Ikaw and the new Ipca) and four ways to aggregate all the information contained in an image (M1, M2 and the new methods M3 and M4) as proposed in [1]. The values are then correlated with SPAD measurements. 

References:

1. Casadesús, Jaume, and Dolors Villegas. "Conventional digital cameras as a tool for assessing leaf area index and biomass for cereal breeding." Journal of integrative plant biology 56, no. 1 (2014): 7-14.

2. Pagola, Miguel, Rubén Ortiz, Ignacio Irigoyen, Humberto Bustince, Edurne Barrenechea, Pedro Aparicio-Tejo, Carmen Lamsfus, and Berta Lasa. "New method to assess barley nitrogen nutrition status based on image colour analysis: comparison with SPAD-502." Computers and electronics in agriculture 65, no. 2 (2009): 213-218

Benefits of the Project

As we know that we are living in a developing country. Most of our peple are facing adverse poverty. Our farmers are very hardworking but at the same time they donot have enough resources and money.The determination of nitrogen status is neccessary to predict crop health which further depends upon chlorophyll amount. If the right ammount is not estimated, it can result in costly overfertilizing or yield-limiting deficiencies. SPAD is being used for measuring the amount of chlorophyll content, But our farmers can not afford this costly instrument. We will develop a smart phone application. As these days every person has access to smart phones and know how to use it. So it will be easily avaible to farmers as it is cost effecient and easy to use. It will help our agricultural industry. And the rise in agricultural production will strengthen our country's economy.

Technical Details of Final Deliverable

We are using computer vision based approach. Our dataset consists of square green leaf images with a black background. In first step we will crop our images to obtain only leaf image. For this we will use mask to detect leaf color through HSV Color space on Opencv with Python. Then we will save the cropped (masked) image. Next, we will calculate a total of eight different greenness indices from digital image. To calculate these, we will use method proposed by Kawashima and Nakatani and the principle component analysis (Ikaw and the new Ipca) as proposed in [1]. Then correlation between green indices and SPAD readings will be determined. After successfully conducting the experiment, we'll write a report at the end.

References:

1. Pagola, Miguel, Rubén Ortiz, Ignacio Irigoyen, Humberto Bustince, Edurne Barrenechea, Pedro Aparicio-Tejo, Carmen Lamsfus, and Berta Lasa. "New method to assess barley nitrogen nutrition status based on image colour analysis: comparison with SPAD-502." Computers and electronics in agriculture 65, no. 2 (2009): 213-218

Final Deliverable of the Project Software SystemCore Industry ITOther Industries Agriculture Core Technology Internet of Things (IoT)Other Technologies Wearables and ImplantablesSustainable Development Goals Decent Work and Economic Growth, Responsible Consumption and ProductionRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	78000
Smart Phone for app development	Equipment	1	45000	45000
Leaf Boards	Equipment	10	300	3000
Field Visits (crops images)	Miscellaneous	5	2000	10000
Chlorophyll Analysis from Agri Labs	Equipment	1	20000	20000