Rice Grain Classification for Agricultural Products Marketing-A Deep Learning Approach

In this research work a deep learning model based on the ResNet50 architecture was developed to categorize five different types of rice: Basmati, Arborio, Jasmine, Ipsala, and Karacadag. A dataset of 2500 rice images, with each class consisting of 500 images has been used. The dataset was split into training and validation sets, with 400 and 100 images per class, respectively. To enhance the model’s robustness and accuracy, the images were pre-processed using data augmentation. The model was trained using the Adam optimizer with a categorical cross-entropy loss. confusion matrix is obtained to evaluate the model’s performance. A comparative analysis of the model with VGG16 and MobileNet was conducted, and the results showed that the proposed model (Resnet50) achieved better accuracy than the others. These findings recommend that deep learning plays a significant role in improving the efficiency and accuracy of rice categorization, which benefits the rice industry by improving quality control, supply chain management, and enhancing consumer confidence in rice products.