Agricultural leaf blight disease segmentation using indices based histogram intensity segmentation approach

Grouping of pixels based on certain kind of similarity or discontinuity among the pixel called Segmentation. Segmentation of ROI from the given input image determines the success of analysis. Validity metrics helps to measure the similarity of the segmented image result. Most important and required for human survival is food. In that scenario Agriculture industry plays a vital role and the industry faces lose because of certain reasons. One of the reason to yield lose is unaware of disease diagnosis and most of the time farmer can predict disease at last moment. By implementing technological improvement in agriculture industry try to improve the crops lose and that results increasing farmer income. Indices based intensity histogram segmentation technique used to segment the disease affected part from unhealthy leaf with better accuracy rate. Segmentation is important stage in image processing technique and it helps to diagnose the diseased region. After categorizing the disease affected area it is most important to validate the segmented image. Validation algorithms are used to validate the segmented part and most famous similarity measures are Dice index measure, over lab coefficient measure, Jaccard coefficient measure, Cosine measure, Asymmetric measure, Dissimilarity measures etc. The introduced method successfully segments the affected region with 98.025% accuracy also the segmented region have 0.964% of mutual information.