Design of Filtration Approach for Image Quality Improvement in Mango Leaf Disease Detection and Pharmaceutical Treatment

Abstract

The traditional method of studying and diagnosing diseases in plants relies primarily on human vision, which is ineffective in identifying diseases in the plants. The color of the leaves gets changed and develops spots such as yellow, brown and black patches as a result of the symptoms. Manually observing leaves for detection is employed to identify the disease, which takes more time, is more costly, and is less accurate. As a result, use of image processing techniques may be a better alternative than certain other old traditional approaches for speedy and exact illness identification. The symptoms may be seen on plant components such as the fruit, leaves, stems, or lesions. The target is to appropriately identify and categorize the diseases based on the leaf photographs. The processes involved in the procedure include image pre-processing, segmentation, feature extraction, and identification. Bacterial, viral, fungal, and insect-borne diseases are all considered. Mango leaves include anthracnose, bacterial canker, and black sooty mold. In this article, a hybrid filter was proposed based on image enhancement i.e., denoising, reducing blurriness and edge sharping of the images and then segmentation done by taking leaves of these three diseases and results are saved.