The intelligent variety identification tool leverages high-resolution images of rice plants at various growth stages. Through sophisticated image processing algorithms and machine learning models. This approach offers a non-destructive, efficient, and accurate means of identifying different rice varieties based on their unique visual characteristics.

Hence, we developed an intelligent tool for identifying rice variety of per plant by deep convolutional neural networks. The training data for the prediction model was WGSR, a dataset containing 93 rice. It offers valuable insights to agricultural researchers, aiding their understanding of rice variety characters and optimizing cultivation management.The benchmark dataset used for this tool, including training and testing sets, is available at opia@big.ac.cn.