Rice Plant Leaf Disease Detection and Severity Estimation

With increase productivity and profit margins, it is imperative to control economic and yield losses of agricultural produce. Manual monitoring of crops is becoming challenging year after year and isn't scalable for large scale cultivation. Hence, in this paper, we discuss various methods used or researched to detect crop diseases in Rice plant using traditional image processing techniques and neural networks. This paper explores possibility of using semantic segmentation to extract the affected area and calculating the affected area and estimate the severity. For easier usage, the model is deployed using ngrok and Twilio server to accept, process and return output on WhatsApp interface.