A Comprehensive Survey of Machine Learning and Deep Learning Methods for Rice Crop Disease Detection

Rice is one of the simple food crops that has been cultivated in the majority of countries. Rice leaf diseases (RLDs) are a major problem in crop production since they may result in low productivity and economic losses. Traditional ways of detecting an illness may be time-consuming and even labor-intensive, and at times may need specialized skills. The popularity of preceding works on detecting RLDs has relied on machine learning (ML) and image processing approaches. On the other hand, deep learning (DL) methodologies are more applicable in disease detection problems because they can learn stipulated patterns on big data without using feature extraction methods. This systematic review explores various ML and DL methods used in the literature for RLD detection, which includes survey articles based on convolutional neural network (CNN), transfer learning, and advanced AI approaches. The review of existing open-source datasets is also discussed in this survey. In addition, it examines limitations of current models related to practical implementation, data diversity, domain adaptation, and hardware limitations. Lastly, this survey identifies future research directions to improve the strength and usage of DL models in real-world agriculture settings. This survey comprehensively reviews more than 70 peer-reviewed publications (2019–2025) sourced from IEEE, Elsevier, Springer, ACM, and MDPI digital libraries.