An Automated Image Processing Method for Segmentation and Quantification of Rust Disease in Maize Leaves

Abstract

$A$ In the agro-ecological domain, maize is one of the most dominantcrops of the world. The disease of the maize plantation not only affect the nutritional balance but also the economics related to the crop. In this paper, an automated image processing method is proposed to identify the rust affected maize leaves and differentiate them from the healthy maize leaves. Segmentation and quantification of the rusted portion from the images of the maize leaf is done. The rust affected portion is accurately quantified of the maize leaf using morphological operations and area based thresholding to make the algorithm computationally efficient. Quantification of the segmented rusted spots is done to measure the degree of damage done by the crop disease. The results obtained from the proposed methodology are encouraging and can be used in agricultural industry for some real-time detection of diseases affecting the productivity of crops.