Enhanced batch sorting and rapid sensory analysis of Mackerel products using YOLOv5s algorithm and CBAM: Validation through TPA, colorimeter, and PLSR analysis

Yi-Zhen Huang; Lin Han; Xiaoqing Yang; Yu Liu; Bei-Wei Zhu; Xiu-Ping Dong

doi:10.1016/j.fochx.2023.100733

Enhanced batch sorting and rapid sensory analysis of Mackerel products using YOLOv5s algorithm and CBAM: Validation through TPA, colorimeter, and PLSR analysis

Food Chem X. 2023 Jun 1:19:100733. doi: 10.1016/j.fochx.2023.100733. eCollection 2023 Oct 30.

Authors

Yi-Zhen Huang^{1

2}, Lin Han^{1

2}, Xiaoqing Yang^{1

2}, Yu Liu^{1

2}, Bei-Wei Zhu^{1

2}, Xiu-Ping Dong^{1

2}

Affiliations

¹ Academy of Food Interdisciplinary Science, School of Food Science and Technology, Dalian Polytechnic University, Dalian 116034, Liaoning, China.
² National Engineering Research Center of Seafood, Collaborative Innovation Center of Seafood Deep Processing, Liaoning Province Collaborative Innovation Center for Marine Food Deep Processing, Dalian 116034, Liaoning, China.

Abstract

This study employed the YOLOv5s algorithm to establish a rapid quality identification model for Pacific chub mackerel (S. japonicus) and Spanish mackerel (S. niphonius). Data augmentation was conducted using the copy-paste augmentation within the YOLOv5s network. Furthermore, a small object detection layer was integrated into the network structure's neck, while the convolutional block attention module (CBAM) was incorporated into the convolutional module to optimize the model. The model's accuracy was assessed through sensory evaluation, texture profile analysis, and colorimeter analysis. The findings indicated that the enhanced model achieved a mAP@0.5 score of 0.966, surpassing the original version's score of 0.953. Moreover, the improved model's params was only 7.848 M, and an average detection time of 115 ms/image (image resolution 2400 × 3200). Furthermore, sensory and physicochemical indicators are reliably distinguished between qualified and unqualified samples. The PLSR model exhibited R²X, R²Y, and Q² values of 0.977, 0.956, and 0.663, respectively.

Keywords: Deep learning; Mackerel; Object detection; Rapid method; Sensory.