An Automated Approach for Plant Health Using Smart-Phone

Abstract

In these days, Smartphone use is very common, almost everyone, including the farmers. The paper presents a scheme that uses smart phones for imaging of diseased plants leaves, proposed work classifies the images of grapes leaves as downy mildew or powdery mildew or healthy and for pomegranate as fungal or bacterial or healthy using SVM, it will gives the percentage amount of the leaf has been infected and remedy suggestion for detected disease. Images of diseased leaves captured by the smartphone camera and the results of processed images will also available on the mobile phone this will achieved by communication between smart phone and local server through internet via matlab application. Hence the system is very much cost effective. The major steps are: image acquisition images are captured by mobile camera, pre-processing step conversion from RGB to L*a*b color space. For multi-thresholding Otsu’s algorithm is used, features are extracted by GLCM technique. The cluster based method is used for segmentation and it is done by k means clustering. For image processing MATLAB software tool is used.