Since farming is the biggest and significant sector of a nation’s economics, it is imperative that new technology be introduced in this area to improve results. Crop health and treatment have a significant impact on yield. To identify illnesses on leaves, a variety of technologies have been developed. In order to identify and treat various illnesses on leaf, this paper will go over on how to segment leaf and its diseased areas. Firstly, the leaf is preprocessed using a canny edge detection to eliminate image’s mask and backdrop, blurring and enhancing the image. Masks are divided into three categories: scab, rust, and healthy. A single multi-mask is created by combining the generated masks with the following values: background = 0, healthy = 1, rust = 2, and scab = 3. Models are now trained and validated using this processed data. Here, segmentation is done using a detection method. It segments sick areas pixel-by-pixel using convolutional and pooling layers. Multiple disease patterns on the leaf can be identified with its help. Every pixel that is a member of one of the predetermined classes is subject to this algorithm. The model has been trained over a set number of epochs with a batch size. As a result, quantitative analysis can be performed to forecast the disease’s severity and the best course of action for curing it.