AI-Driven Rice Leaf Nutrient Deficiency Detection Using Deep Learning with Explainable Insights

Abstract

Rice is eaten every day, and it’s an important staple. The production of rice depends upon the availability of nitrogen, phosphorus, potassium and the health of the plants. The health of the rice plant and the production of rice depend on knowing whether the rice plant is low in certain substances. The conventional procedure entails inspecting the plants and laboratory testing, which is time-consuming and requires expertise. The purpose of the present study is to design a system that can identify nutrient deficiency on rice leaves using specialized software programs in an automatic manner. The system uses a computational model that is good at looking at pictures and figuring out what is in them. We resize the photos of the rice leaves to a standard size before feeding them to the computer model. The computer model is able to identify if the leaf is healthy or unhealthy, and how sure it is about this evaluation. The method also points out the exact parts of the leaf that are affected and explains why the computer model made its decision. The system can evaluate the severity of the problem and suggest the right fertilizer to use, how much it will cost and when to apply it. The use of these computer tools and providing access to the farmers is an effective approach to early diagnosis and solution of rice leaf problems. Rice is an agricultural product, and this technique can boost our cultivation of rice. The technology is easy to use and helps farmers control rice production, and is advantageous to farmers.