DETECTION OF DISEASES ON LEAVES AND ITS POSSIBLE DIAGNOSIS USING CBIR TECHNIQUE

Image Retrieval has been an important achievement of technological advancement in today's Age. In the process of retrieval, the features viz. colour, texture, entropy, densities are brought in a consideration for classification of images. The range of scope varies among various applications and proves to be useful for agricultural sector. This paper proposes the detection of diseases on the leaves of plants depicting radical changes over the surface (e.g. cotton plant). The difference between a normal cotton plant's leaf and an infected one is visually forecasted based on size, texture, colour, and density. As previous Methodologies have shown lack of processing abilities to a large database, CBIR technique is proposed for handling of large operations. CBIR technique is provided out as a combination of content based retrieval, colour analysis and data-mining techniques. Thus, following information is used to segment leaf image into pixel format. The software analysis is further based upon the nature of image, producing respective results.