A rapid identification technique for rice adulteration based on improved YOLOV8 model

Abstract Rice classification and quality testing are essential to ensure its safety and quality and are effective in reducing rice-related food economy losses.Currently, incidents of rice adulteration have been repeatedly reported.For this reason, this paper optimises and proposes the lightweight and efficient Faster-YOLO algorithm on the basis of YOLOv8n algorithm, which is more suitable for the task of rice adulteration classification and recognition. Firstly, this paper introduces the grouped convolutional hybrid attention mechanism (GCHAM) combining channel information and spatial information, which is embedded in the last layer of the backbone network to enhance the model feature representation capability by focusing on the key information in order to suppress the noise. Secondly, the C2F module in the backbone part adopts the design of combining Faster and C2F to enhance the feature fusion capability and reduce the weight of the model, thus reducing the number of parameters and FLOPs.Finally, the collected data are augmented with multiple aspects to simulate different environments, and compared with multiple attention mechanisms and deep learning models. The experimental results show that the proposed method in this paper is superior in classification and recognition performance, with recognition accuracy of 93.4%, precision of 93.4%, recall of 93.6%, and F1 score of 93.5%. It proves that Faster-YOLO improves the detection and recognition ability while reducing the weight of the model, which provides a strong basis for the rapid identification of rice adulteration.