Potato Plant Leaf Disease Classification Using Deep CNN

The agricultural sector is a significant contributor to the national economy. There is nothing more fundamental to a country than this. Unfortunately, agricultural products are more vulnerable to various diseases. Potato is a crucial component of the diet of 1.5 billion people worldwide and the primary non-cereal agricultural product in the world food chain. One potato seed can produce about ten tubers of potatoes. Potato diseases are therefore common since they spread quickly through seed potatoes. As a result, identifying diseases in potato plants at an early stage becomes critical. This also encourages the development of automated techniques for identifying diseases in potato plants. Machine learning and deep learning have made significant advances in the development of automated systems for disease detection and classification in recent years. Various researchers have worked in this area and developed algorithms for the same. In this paper we applied different deep CNN model like VGG19, ResNet50, DenseNet121 to detect potato disease from its leaves. We also developed our shallow deep CNN model to provide more accurate results and it gives accuracy 94.11% and it is better than other deep CNN models.