Damage Detection of X-ray Image of Conveyor Belts with Steel Rope Cores Based on Improved FCOS Algorithm

doi:10.1007/s12204-023-2651-6

Abstract

Abstract: Aimed at the long and narrow geometric features and poor generalization ability of the damage detection in conveyor belts with steel rope cores using the X-ray image, a detection method of damage X-ray image is proposed based on the improved fully convolutional one-stage object detection (FCOS) algorithm. The regression performance of bounding boxes was optimized by introducing the complete intersection over union loss function into the improved algorithm. The feature fusion network structure is modified by adding adaptive fusion paths to the feature fusion network structure, which makes full use of the features of accurate localization and semantics of multi-scale feature fusion networks. Finally, the network structure was trained and validated by using the X-ray image dataset of damages in conveyor belts with steel rope cores provided by a flaw detection equipment manufacturer. In addition, the data enhancement methods such as rotating, mirroring, and scaling, were employed to enrich the image dataset so that the model is adequately trained. Experimental results showed that the improved FCOS algorithm promoted the precision rate and the recall rate by 20.9% and 14.8% respectively, compared with the original algorithm. Meanwhile, compared with Fast R-CNN, Faster R-CNN, SSD, and YOLOv3, the improved FCOS algorithm has obvious advantages; detection precision rate and recall rate of the modified network reached 95.8% and 97.0% respectively. Furthermore, it demonstrated a higher detection accuracy without affecting the speed. The results of this work have some reference significance for the automatic identification and detection of steel core conveyor belt damage.

Key words: conveyer belts with steel rope cores, damage, X-ray image, image detection, improved fully convolutional one-stage object detection (FCOS) algorithm

摘要： 针对钢丝绳芯输送带损伤X射线图像识别中识别目标长、窄几何特征和泛化能力较差等问题，提出一种基于改进FCOS算法的钢丝绳芯输送带损伤X射线图像识别方法。在改进算法中采用CIoU损失函数优化包围框回归性能；在特征融合网络结构中添加自适应融合路径，充分利用多尺度特征融合网络定位精准、语义性强的特点，提出新的特征融合网络结构。最后，利用某探伤装置公司提供的钢丝绳芯输送带损伤X射线图像数据集进行网络训练和验证，并采用镜像翻转、旋转以及缩放等数据增强方法丰富钢丝绳芯输送带损伤特征图像数据集，以保证模型得以充分训练。实验结果表明，改进算法在原算法基础上查准率提高20.9%，查全率提高14.8%。同时与Fast R-CNN、Faster R-CNN、SSD、YOLOv3网络进行比较，优势较为明显，查准率和查全率分别达到95.8%、97.0%，且在检测速度不影响的情况下，有更高的检测精度。该研究结果对钢丝绳芯输送带损伤的自动识别和检测有重要参考意义。

关键词: 钢丝绳芯输送带，损伤，X射线图像，图像检测，改进FCOS算法

CLC Number:

Wang Baomin, Ding Hewei, Teng Fei, Liu Hongqin. Damage Detection of X-ray Image of Conveyor Belts with Steel Rope Cores Based on Improved FCOS Algorithm[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(2): 309-318.

References

[1] LI S, XUE G, FANG X, et al. Coal mine intelligent safety system and key technologies [J]. Journal of China Coal Society, 2020, 45(6): 2320-2330 (in Chinese).
[2] LONG X Y, LI X G, SUN M W, et al. Quantitative analysis of bond and splice strength of steel cord conveyor belt [J]. Journal of Adhesion Science and Technology, 2020, 34(14): 1544-1555.
[3] WANG C Q, ZHANG J A. The research on the monitoring system for conveyor belt based on pattern recognition [J]. Advanced Materials Research, 2012, 466/467: 622-625.
[4] TIAN S C, LI H X, WANG L. Three types hazard theory and prevention of coalmine accidents [J]. Journal of China Coal Society, 2006, 31(6): 706-710 (in Chinese).
[5] XU K, ZHOU P, YANG Z L. On-line detection technique of tiny surface defects for metal plates and strips based on photometric stereo [J]. Journal of Mechanical Engineering, 2013, 49(4): 25-29 (in Chinese).
[6] PENG C, LIU B, ZHOU Q. Mechanical fault detection based on machine vision and blind source separation [J]. Journal of Shanghai Jiao Tong University, 2020, 54(9): 953-960 (in Chinese).
[7] GUAN Y, ZHANG J, SHANG Y Y, et al. Embedded sensor of forecast conveyer belt breaks [C]//2008 Fifth International Conference on Fuzzy Systems and Knowledge Discovery. Jinan: IEEE, 2008: 617-621.
[8] BERNIERI A, FERRIGNO L, LARACCA M, et al. Ultrasonic NDT on aluminum bars: An experimental performance comparison of excitation and processing techniques [J]. Measurement, 2018, 128: 393-402.
[9] HU X Z. Penetration testing [M]. Beijing: The China Labor and Social Security Publishing House, 2007 (in Chinese).
[10] Nondestructive Testing Branch of the Chinese Society of Mechanical Engineering. Magnetic particle testing [M]. Beijing: Machinery Industry Press, 2004 (in Chinese).
[11] REDKO V, KHANDETSKYY V, NOVAK P, et al. Method and apparatus for eddy current-based quality inspection of dry electrode structure: US7355395 [P]. 2008-04-08.
[12] BLAZEJ R, JURDZIAK L, KOZWSKI T, et al. The use of magnetic sensors in monitoring the condition of the core in steel cord conveyor belts - Tests of the measuring probe and the design of the DiagBelt system [J]. Measurement, 2018, 123: 48-53.
[13] QUAN D Y, CHEN Y, TANG Z Y, et al. Radar signal recognition based on dual channel convolutional neural network [J]. Journal of Shanghai Jiao Tong University, 2022, 56(7): 877-885 (in Chinese).
[14] XU Z Q, L¨U Z Q, WANG W D, et al. Machine vision recognition method and optimization for intelligent separation of coal and gangue [J]. Journal of China Coal Society, 2020, 45(6): 2207-2216 (in Chinese).
[15] ALPORT M, GOVINDER P, PLUM S, et al. Identification of conveyor belt splices and damages using neural networks [J]. Bulk Solids Handling, 2001, 21(6): 622-627.
[16] FOURIE J, ALPORT M, BASSON J, et al. Condition monitoring of fabric-reinforced conveyor belting using digital x-ray imaging [J]. Bulk Solids Handling, 2005, 25(5): 290-294.
[17] WANG J F, MIAO C Y, WANG W, et al. Research of X-ray nondestructive detector for high-speed running conveyor belt with steel wire ropes [J]//Proceedings of SPIE, 2007, 6833: 482-490.
[18] WU S C, QI Z F, LI J X. Intelligent global sensitivity analysis based on deep learning [J]. Journal of Shanghai Jiao Tong University, 2022, 56(7): 840-849 (in Chinese).
[19] ZHANG M C, SHI H, ZHANG Y, et al. Deep learningbased damage detection of mining conveyor belt [J]. Measurement, 2021, 175: 109130.
[20] HU Z, XU C Z, LEI G B, et al. Real-time detection of casting defect in X-ray images based on improved YOLOv5 [J]. Laser Journal, 2022, 43(5): 6 (in Chinese).
[21] HAO S, ZHANG X, MA X, et al. Foreign object detection in coal mine conveyor belt based on CBAMYOLOv5 [J]. Journal of China Coal Society, 2022, 47(11): 4147-4156 (in Chinese).
[22] ZHANG X L, DONG X P, WEI Q J, et al. Realtime object detection algorithm based on improved YOLOv3 [J]. Journal of Electronic Imaging, 2019, 28(5): 053022.
[23] LIU W M, XIAO Y, ZHENG A Y, et al. Research on fault diagnosis of steel surface based on improved YOLOV5 [J]. Processes, 2022, 10(11): 2274.
[24] WU D B, LIU X N. SSD small target detection method based on multi-scale semantic information fusion [J]. Microprocessors, 2022, 43(3): 51-55 (in Chinese).
[25] PANDEY A K, BISWAS M, SAMMAN M M. Damage detection from changes in curvature mode shapes [J]. Journal of Sound and Vibration, 1991, 145(2): 321- 332.
[26] GE Y Z, LIU H, WANG Y, et al. Survey on deep learning image recognition in dilemma of small samples [J]. Journal of Software, 2022, 33(1): 193-210 (in Chinese).
[27] LI J H, LIN L J, TIAN K, et al. Detection of leaf diseases of balsam pear in the field based on improved Faster R-CNN [J]. Transactions of the Chinese Society of Agricultural Engineering, 2020, 36(12): 179-185 (in Chinese).
[28] SHU J Q, XU Y H, XIA T B, et al. A multiscale similarity ensemble methodology for remaining useful life prediction in multiple fault modes [J]. Journal of Shanghai Jiao Tong University, 2022, 56(5): 564-575 (in Chinese).
[29] TIAN Z, SHEN C H, CHEN H, et al. FCOS: fully convolutional one-stage object detection [C]//2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 9626-9635.
[30] GHIASI G, LIN T Y, LE Q V. NAS-FPN: Learning scalable feature pyramid architecture for object detection [C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach: IEEE, 2019: 7029-7038.
[31] LI M, LU P, ZHU M Q, et al. Opening degree detection of gate valve based on improved YOLO-tiny [J]. Journal of China Coal Society, 2021, 46(S2): 1180-1190 (in Chinese).
[32] ZHENG Z H, WANG P, LIU W, et al. Distance-IoU loss: Faster and better learning for bounding box regression [J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2020, 34(7): 12993-13000.
[33] LIU S, QI L, QIN H F, et al. Path aggregation network for instance segmentation [C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 8759-8768.