17 May 2022 in Measurement Science and Technology
This work proposes a recognition method to improve the visual detection rate of longitudinal conveyor belt tears related to fog pollution and uneven illumination. It integrates the Haar feature with a modified dark channel defogging algorithm. Firstly, the dark channel algorithm was used to eliminate fog interference with belt images, and the processing time was reduced from 14.5 ms to 5 ms by calculating a single necessary image channel. The Haar feature was used to replace the traditional geometric features due to its high-efficiency feature expression and illumination robustness; this provided a strong feature basis for the learning and training of the classifiers. Next, the classifiers were trained and strengthened by the AdaBoost algorithm, and the strengthened classifiers were cascaded by the Cascade algorithm to improve the recognition ability. Finally, experiments were carried out under mixed fog pollution and uneven illumination. The results show that the average accuracy, recall, precision and false positive rate of the proposed method are approximately 98.4%, 96.9%, 99.9% and 0.1%, respectively, while the recognition period was only 52.3 ms. This shows strong adaptability and robustness to a harsh environment and demonstrates that the proposed method has important application value.
14 July 2023 in Journal of Failure Analysis and Prevention
28 February 2024 in Journal of Failure Analysis and Prevention
1 June 2024 in Engineering Research Express
In the industrial production process, the detection of conveyor belt damage plays a crucial role in ensuring the stable operation of the transportation system. To tackle the issues of significant changes in damage size, missed detections, and poor detection ability of small-size objects in conveyor belt surface damage detection, an improved HLG-YOLOv7 (Hybrid Local and Global Features Network) conveyor belt surface defect detection algorithm is proposed. Firstly, Next-VIT is employed as the backbone network to fully extract local and global features of the damage, enhancing the model’s ability to extract features of different-sized damages. Additionally, to deeply utilize the extracted local and global features, the Explicit Visual Center (EVC) feature fusion module is introduced to obtain comprehensive and discriminative feature representations, further enhancing the detection capability of small objects. Lastly, a lightweight neck structure is designed using GSConv to reduce the complexity of the model. Experimental results demonstrate that the proposed method has better small object detection performance compared to existing methods. The improved algorithm achieves mAP and F1 scores of 96.24% and 97.15%, respectively, with an FPS of 28.2.