Identification And Classification of Rice Leaf Disease Using Hybrid Deep Learning

Abstract

Agriculture is the primary source of income and livelihood in India as well as in many countries. Rice crop is considered one of the most cultivated grain crops in India. Rice crop is most essential crop for human consumption. Around half of the global population relies on cereal as a primary food source. Now a days this crop is susceptible to various types of illnesses at different stages of production, this can be effect yield and quality of rice crops. As a result, automation of identification and early diagnosis of rice leaf disease is widely needed in the agriculture field. Using CNN-VGG algorithm the suggested system has been used to identify disease in rice crop. In this study we focus on three well know rice leaf disease such as brown spot, leaf blast and leaf blight, the total dataset of 1800 images of 4 classes are taken. When compared to existing model our experimental result analysis of CNN with Transfer learning has a higher accuracy of 95.60%, and the same data set is also applied to CNN without transfer learning has achieved an 85.62%. The proposed model attained a higher accuracy.