The key factors that determines the loss of agricultural and crop production output is to recognize plant diseases. Plant disease research focuses on any apparent features, such as spots or color variations that enable us to distinguish between two different types of plants. The important factors in the growth of agriculture is the sustainability of plants. Accurately identifying plant diseases is quite challenging. A great deal of effort and skill to know the disease, extensive knowledge of plants and research on disease detection. When detecting illnesses, the procedures of image capture, extraction, segmentation, and pre-processing are used. Various illnesses affect the amount of chlorophyll in leaves, resulting in brown or black spots on the leaf’s surface. The economy has a big impact on agricultural productivity. The farmer faces many obstacles when they switch between various disease management methods. Classification uses CNN, a kind of deep learning technique. Transformer networks have recently showed a lot of potential in computer vision problems. In order to detect plant diseases, this study contrasts these methods with conventional CNN methods. Our transformer model’s highest validation accuracy is 97.98%.