Color segmentation to measure the percentage of the affected area in leaves with signs of chlorosis

Abstract

Observing the signs of deterioration caused by environmental pollutants and some phytopathogens in plants, a computational algorithm programmed in Python language was developed using image processing tools to determine the percentage of damage in leaves with signs of chlorosis. In the present study, the following stages were implemented, i) image collection, using bean (Phaseolus vulgaris) plants exposed to perchlorate, and cowpea (Vigna unguiculata) plants affected by the phytopathogenic fungus Rhizoctonia solani AG-1 IA. ii) Image processing, by implementing the OpenCV-Python package that allowed segmentation and binarization of the images. Finally, the result of the binarization was compared with an approximation of a healthy leaf, and the percentage of affected leaf area compared with the healthy leaf obtained. Meanwhile, the timely detection of diseases in plants and crops is a determining factor for the efficiency of agricultural production, as well as the assessment and presence of chemical substances that affect the environment and human health.