In the traditional operation mode, the quantity verification work when the vehicle transport coils steel in and out of the logistics gate is completed by the gate inspectors, which is labor-intensive and costly, and cannot meet the requirements of modern logistics enterprises. In order to solve this problem, a vehicle coil wire counting method for high-shot viewing angles of unattended gates is designed using target recognition technology. Firstly, the improved YOLOv7 model performs three recognition rounds on vehicle-mounted circular images from high camera angles, classify their stacking forms, and detects them. Secondly, a design calculation algorithm is proposed based on different stacking forms to calculate the number of circles automatically. Applying the method proposed in this paper to the detect real gates and optimize them for practicality. By selecting an appropriate target detection confidence threshold, the accuracy of the exit gate was ensured to be at a high level. Then, through the three-attempt method, the intercepted vehicle was guided to adjust its position to improve the accuracy of the intercepted vehicle and the exit gate rate of the legitimate vehicle. The experimental results showed that the improved YOLOv7 model effectively enhanced the effectiveness of circle detection; the proposed method for calculating the number of discs has an accuracy rate of over 90%. Following practical optimization efforts, the counting accuracy of on-board disc steel has been enhanced to 97.88%, and the exit rate of legitimate vehicles within three inspections has risen to 98.27%. These achievements meet the usability standards for port on-site operations.
BELMOUKARIB, AUDYJ F, FORGETP. Smart port: A systematic literature review[J]. European Transport Research Review, 2023, 15(1): 4. DOI: 10.1186/s12544-023-00581-6 .
CAOL G, MIC. Research on recognition algorithms for container code character[J]. Computer Engineering and Applications, 2021, 57(15): 178-185. DOI: 10.3778/j.issn.1002-8331.2004-0254(Ch ).
[4]
GIRSHICKR, DONAHUEJ, DARRELLT, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE Press,2014: 580-587. DOI: 10.1109/CVPR.2014.81 .
[5]
RENS Q, HEK M, GIRSHICKR, et al. Faster R-CNN: Towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149. DOI: 10.1109/TPAMI.2016.2577031 .
[6]
REDMONJ, DIVVALAS, GIRSHICKR, et al. You only look once: Unified, real-time object detection[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). New York: IEEE Press, 2016: 779-788. DOI: 10.1109/CVPR.2016.91 .
[7]
LIUW, ANGUELOVD, ERHAND, et al. SSD: Single shot MultiBox detector[C]//European Conference on Computer Vision. Cham: Springer, 2016: 21-37.10.1007/978-3-319-46448-0_2. DOI: 10.1007/978-3-319-46448-0_2 .
WANGC, ZHANGY C, SUNS H, et al. Steel-bar end face detection based on improved YOLOv5 algorithm[J]. Computer Systems and Applications, 2022, 31(4): 68-80. DOI: 10.15888/j.cnki.csa.008433(Ch ).
XIEH Z. Research and application of rebar counting algorithm in complex scene[D].Chengdu: University of Electronic Science and Technology of China, 2020(Ch).
[14]
WANGC Y, BOCHKOVSKIYA, LIAOH Y M. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[EB/OL]. 2022: arXiv: 2207.02696. DOI: 10.1109/cvpr52729.2023.00721 .
[15]
QINX Y, LIN, WENGC, et al. Simple attention module based speaker verification with iterative noisy label detection[C]//ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). New York: IEEE Press, 2022: 6722-6726. DOI: 10.1109/ICASSP43922.2022.9746294 .
[16]
DINGX H, ZHANGX Y, MAN N, et al. RepVGG: Making VGG-style ConvNets great again[C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New York: IEEE Press, 2021: 13728-13737. DOI: 10.1109/CVPR46437.2021.01352 .