Intelligent classification and identification method for Conger myriaster freshness based on DWG‐YOLOv8 network model

• Published in: Food bioengineering Vol. 3; no. 3; pp. 269 - 279
• Main Authors: Gao, Sheng, Wang, Wei, Lv, Yuanmeng, Chen, Chenghua, Xie, Wancui
• Format: Journal Article
• Language: English
• Published: Wiley 01.09.2024
• Subjects: classification recognition; computer vision; Conger myriaster freshness; deep learning; DWG‐YOLOv8 network
• ISSN: 2770-2081; 2770-2081
• DOI: 10.1002/fbe2.12097

The freshness of aquatic products is directly related to the safety and health of the people. Traditional methods of detecting the freshness of Conger myriaster rely on manual operations, which are labor‐intensive, inefficient, and highly subjective. This paper combines computer vision and the DWG‐YOLOv8 network model to establish an intelligent classification method for C. myriaster freshness. Through image augmentation, 484 C. myriaster samples were expanded to 2904 samples. The YOLOv8n model was improved by simplifying the network backbone, introducing Ghost convolution and the new DW‐GhostConv, thereby reducing the number of parameters and computational load. Test results show that the recognition accuracy of the DWG‐YOLOv8 model reached 98.958%, outperforming models such as ResNet18, Mobilenetv3 small, and Swin transformer v2 tiny. The model's parameter count is 16.609 K, the inference time is 57.80 ms, and the model size is only 102 KB. The research provides a reliable method for online intelligent and nondestructive detection of C. myriaster freshness. This study proposes an intelligent classification and identification method based on an improved DWG‐YOLOv8 deep learning network model for evaluating the freshness of Conger myriaster. The graphical illustrates the overall workflow of the proposed method, including image data acquisition, model construction and optimization, and the final freshness detection process. By incorporating Ghost convolution and the newly developed DW‐GhostConv module into the YOLOv8n model, this study simplifies the network structure while maintaining high recognition accuracy. The improved DWG‐YOLOv8 model demonstrated excellent recognition accuracy in testing. Additionally, the model has a low computational load and a compact size, making it particularly suitable for online, non‐destructive detection of Conger myriaster freshness.