Publication

Abstract : Rice, as a primary dietary component for most of the world's population, assumes critical significance in ensuring global food security. Accurately classifying rice image is a crucial first step in promoting effective agricultural production and supporting initiatives to ensure food security. This study is a compelling testament to the potential of deep learning in rice image classification, accentuating its pivotal role as a valuable tool for augmenting agricultural productivity, ensuring food availability, and contributing substantively to global food security endeavours. Proposed method uses a painstakingly chosen rice dataset to optimize the pre-trained VGG16 and MobileNetv2 models. The categorization results achieved with the VGG16 and MobileNetv2 models are closely examined. Notably, the MobileNetv2 and VGG16 architecture shows an outstanding accuracy of 99.5%.