A Novel AI-Based Approach for Better Segmentation of the Fungal and Bacterial Leaf Diseases of Rice Plant

Rice is the most consumed food for more than half world. All over world, approximately 15% of rice get wasted because leaf diseases. A computer-aided system needs a clear segmented lesion to detect such diseases, but blurriness, bad contrast, and dust particles on leaves are challenge in proper segmentation and, further, better feature extraction. In this work, first, SegNet deep learning model was trained separate weed from images captured field; then, next step, novel automated technique named RPK-means proposed combining random path (RP) K-means clustering spots images. The work multifold. First, separation; two clusters image generated by K -means find out pixel coordinates lying spot healthy part leaves. Thereafter, background, automatic performed id="M2"> (RPK-means) method using coordinate positions obtained at last stage. Fungal bacterial diseases like brown spot, blast, sheath blight, scaled, blight have been collected field perform experiments. Experimental result shows that performance classifier increased 2-6% while applying preprocessed images, rather traditional id="M3"> technique.