Plant Disease Recognition and Crop Recommendation System using Deep Learning

Agriculture is the foundation of our country. India, the world's second-largest producer of agricultural goods, generates over 280 million tonnes annually, which accounts for more than 15% of its GDP. A.A.R.C.S. (Advanced Agricultural Recommendation and Classification System) is a flutter-based farmer-friendly Mobile Application. One of the features of the proposed system is the crop recommendation feature that recommends correct crop by considering parameters like potassium, phosphorous and nitrogen contents, temperature, rainfall, and humidity. The second feature is seedling classification (using transfer learning), to identify the plants at an early stage. And as the third feature, the system also uses transfer learning to teach the model to detect common plant diseases. The above two models accept the image as an input and recognize and classify the image into the corresponding class. The proposed system uses edge-computing technology (Deep-Learning based model inside the mobile application) which simply means “taking action on data as near to the source as possible rather than in a central, remote data center, to reduce latency and bandwidth use”, i.e. the classification will be done on-device itself, thus overcoming the need of high latency and high bandwidth network links. The application also features a common platform to list the products of the farmers and tools and lands for rent service by providing the respective contact numbers of the owners. There will also be an encrypted community chat option which will be implemented using network sockets. The proposed system also contains an admin dashboard that provides infographics about users, common diseases, chat usage, push notification service and its analytics, etc. The mobile application will be the data source for disease-based analytics.