Deep Transfer Learning-Based Rice Leaves Disease Diagnosis and Classification model using InceptionV3
Rukhsar, Santosh Kumar Upadhyay
Abstract
Rice (Oryza sativa) is the globe's favorite eatable grain. It is the important diet for more than half of the globe's population as a source of energy. Abiotic and biotic elements such as weather, soil fertility, temperature, pests, pathogens, viruses, and others influence rice grain yields production amount and quality. Farmers invest a lot of time and energy in disease control, and they detect diseases with their penniless human eye strategy, which causes unhealthy cultivation. To avoid biased, incorrect, and inefficient manual detection, we are introducing an automatic plant disease recognition and classification method by using a transfer learning technique. The deep neural networks algorithm is a convolutional learning technique that has been successfully used for computer vision problems like as image processing, object identification, and image identification. InceptionV3 is a type of CNN architecture used in our research to recognize diseases in rice leaf images using a transfer learning approach. The developed model's results are obtained for the classification problem, and it achieved a high accuracy of 90.77 percent.