LAUSR

Abstract Crop diseases in India are a major concern in the field of agriculture today. Efforts are being made to tackle this problem as agriculture is an important aspect of our economy. Due to diseases, a major part of crops produced is lost every year. This is due to lack of expertise among the farmers, failing to detect and diagnose the disease correctly in time or due to commencement of control measure in late. To prevent crop loss, image processing for diagnosis of disease symptom using soft computing techniques is a need of modern day agriculture. In this study, initial focus was paid towards automatic identification of leaves between okra and bitter gourd from the images. To diagnose disease symptom from leaf images, forty-three morphological features of leaves of both the crops were chosen. The Pearson Correlation Coefficient technique was exercised to set key identifying features depending on a particular type of leaf image. Ten key morphological features of okra leaf and nine key morphological features of bitter gourd leaf were selected for the purpose. The proposed technique was experimented on 79 and 75 okra and bitter gourd leaf images, respectively. Finally Yellow Vein Mosaic Virus (YVMV) disease gradation was done based on entropy based binning and Naive Bayes classifier. The success rate of leaf identification was 96.78%.YVMV disease affected leaves were graded with a success rate of 95% for okra and 82.67% for bitter gourd.