改进YOLO11n和PaddleOCR的煤矿钻场视频自动剪辑方法

doi:10.3778/j.issn.1002-8331.2503-0187

摘要/Abstract

摘要： 为解决煤矿井下瓦斯抽采钻场监控视频数据规模大、传统人工剪辑效率低的问题，提出一种将YOLO11n和PaddleOCR相结合的视频自动剪辑方法。使用YOLO11n检测视频图像帧中的指示牌目标，并根据检测框坐标信息进行裁剪；将裁剪的目标区域输入PaddleOCR中进行文字识别；依据设定的剪辑逻辑规则对视频进行自动剪辑。为提升YOLO11n在煤矿井下复杂环境的检测精度，提出一种新的模块Faster-EMA来替代C3k2中的Bottleneck，引入FasterBlock及EMA注意力机制，增强多尺度特征表达能力并降低冗余计算；在C2PSA层后引入Triplet Attention，通过三分支结构捕获跨维交互来计算注意力权重，进一步增强特征提取效果；采用PIoUv2替代默认损失函数CIoU以解决锚框扩展问题。同时使用改进后的YOLO11n替换PaddleOCR中的文本检测算法DBNet，解决实时性不足问题。在自建的指示牌数据集上进行实验验证，结果表明，改进的YOLO11n对比原模型，mAP50提升4.8个百分点，且使用改进YOLO11n替代DBNet后视频平均处理速度提升51.0%，FPS达到37帧/s，满足实时性需求。研究实现了基于指示牌文字内容的钻场监控视频自动剪辑，为煤矿智能化发展提供了技术参考。

关键词: 煤矿钻场, 视频剪辑, 文本识别, YOLO11n, PaddleOCR

Abstract: A video automatic editing method combining YOLO11n and PaddleOCR is proposed to solve the problems of large scale of monitoring video data and low efficiency of traditional manual editing in underground gas extraction drilling sites in coal mines. This paper first uses YOLO11n to detect signage targets in video image frames and crops them based on the coordinate information of the predicted box. Next, it inputs the cropped target area into PaddleOCR for text recognition. Finally, it automatically edits the video according to the set editing logic rules. To improve the detection accuracy of YOLO11n, a new module Faster-EMA is proposed to replace Bottleneck in C3k2. This module introduces FasterBlock and EMA attention mechanism to enhance multi-scale feature expression ability and reduce redundant computation. It introduces Triplet Attention after the C2PSA layer, captures cross dimensional interactions through a three branch structure to calculate attention weights and further enhances feature extraction performance. It uses PIoUv2 instead of the default loss function CIoU to solve the anchor box extension problem, and to speed up video processing, and it uses the improved YOLO11n to replace the text detection algorithm DBNet in PaddleOCR. Experimental verification is conducted on the self-built signage dataset, and the results show that the improved YOLO11n model improves mAP50 by 4.8 percentage points compared to the original model. Moreover, the use of the improved YOLO11n instead of DBNet increases the average video processing speed by 51.0%, and the FPS reaches 37 frames per second, meeting real-time requirements. This paper has achieved automatic video editing of drilling site monitoring based on signage text content, providing reference and technical support for the intelligent development of coal mines.

Key words: coal mine drilling sites, video editing, text recognition, YOLO11n, PaddleOCR

李小军, 李淼, 赵明炀. 改进YOLO11n和PaddleOCR的煤矿钻场视频自动剪辑方法[J]. 计算机工程与应用, 2025, 61(17): 209-221.

LI Xiaojun, LI Miao, ZHAO Mingyang. Automatic Video Editing Method for Coal Mine Drilling Site Based on Improved YOLO11n and PaddleOCR[J]. Computer Engineering and Applications, 2025, 61(17): 209-221.

参考文献

[1] 荣耀, 安晓宇. 智能化开采中视频信息的应用现状及展望[J]. 煤炭科学技术, 2021, 49(S1): 119-123.
RONG Y, AN X Y. Application status and prospect of video information in intelligent mining[J]. Coal Science and Technology, 2021, 49(S1): 119-123.
[2] 王国法, 任怀伟, 富佳兴. 煤矿智能化建设高质量发展难题与路径[J]. 煤炭科学技术, 2025, 53(1): 1-18.
WANG G F, REN H W, FU J X. The challenge and path of high-quality development of coal mine intelligent construction[J]. Coal Science and Technology, 2025, 53(1): 1-18.
[3] 王国法, 庞义辉, 任怀伟, 等. 智慧矿山系统工程及关键技术研究与实践[J]. 煤炭学报, 2024, 49(1): 181-202.
WANG G F, PANG Y H, REN H W, et al. System engineering and key technologies research and practice of smart mine[J]. Journal of China Coal Society, 2024, 49(1): 181-202.
[4] 刘艳菊, 伊鑫海, 李炎阁, 等. 深度学习在场景文字识别技术中的应用综述[J]. 计算机工程与应用, 2022, 58(4): 52-63.
LIU Y J, YI X H, LI Y G, et al. Application of scene text recognition technology based on deep learning: a survey[J]. Computer Engineering and Applications, 2022, 58(4): 52-63.
[5] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2016: 779-788.
[6] REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 6517-6525.
[7] REDMON J, FARHADI A. YOLOv3: an incremental improvement[J]. arXiv:1804.02767, 2018.
[8] DU Y N, LI C X, GUO R Y, et al. PP-OCR: a practical ultra lightweight OCR system[J]. arXiv:2009.09941, 2020.
[9] DU Y N, LI C X, GUO R Y, et al. PP-OCRv2: bag of tricks for ultra lightweight OCR system[J]. arXiv:2109.03144, 2021.
[10] 高如新, 杜亚博, 常嘉浩. 基于改进YOLOX-S的轻量化煤矸石检测方法研究[J]. 河南理工大学学报(自然科学版), 2024, 43(4): 133-140.
GAO R X, DU Y B, CHANG J H. Study on lightweight coal gangue detection method based on improved YOLOX-S[J]. Journal of Henan Polytechnic University (Natural Science), 2024, 43(4): 133-140.
[11] 徐壮, 钱育蓉, 颜丰. GCW-YOLOv8n: 轻量级安全帽佩戴检测算法[J]. 计算机工程与应用, 2025, 61(3): 144-154.
XU Z, QIAN Y R, YAN F. GCW-YOLOv8n: lightweight safety helmet wearing detection algorithm[J]. Computer Engineering and Applications, 2025, 61(3): 144-154.
[12] 王琦, 夏鲁飞, 陈天明, 等. 基于改进YOLOv8n的井下人员安全帽佩戴检测[J]. 工矿自动化, 2024, 50(9): 124-129.
WANG Q, XIA L F, CHEN T M, et al. Detection of underground personnel safety helmet wearing based on improved YOLOv8n[J]. Journal of Mine Automation, 2024, 50(9): 124-129.
[13] 张娜, 彭文韬, 赵强, 等. 基于MBE-YOLO的轻量化井下行人检测模型[J]. 矿业研究与开发, 2025, 45(5): 192-199.
ZHANG N, PENG W T, ZHAO Q, et al. Lightweight underground pedestrian detection model based on MBE-YOLO[J]. Mining Research and Development, 2025, 45(5): 192-199.
[14] 姜媛媛, 刘宋波. 基于改进YOLOv8n的煤矿井下钻杆计数方法[J]. 工矿自动化, 2024, 50(8): 112-119.
JIANG Y Y, LIU S B. A coal mine underground drill pipes counting method based on improved YOLOv8n[J]. Journal of Mine Automation, 2024, 50(8): 112-119.
[15] 瞿鹏程, 李敬兆, 刘泽朝. 基于改进YOLOv7和Byte Track的煤矿多目标人员跟踪算法[J]. 煤矿安全, 2025, 56(1): 195-205.
QU P C, LI J Z, LIU Z C. Multi-target personnel tracking algorithm for coal mine based on improved YOLOv7 and Byte Track[J]. Safety in Coal Mines, 2025, 56(1): 195-205.
[16] 李志星, 杨啸龙, 李天昊, 等. 基于YOLOv8的煤矿用钢丝绳损伤检测算法[J]. 电子测量技术, 2024, 47(9): 120-128.
LI Z X, YANG X L, LI T H, et al. YOLOv8-based surface damage detection of mine wire rope[J]. Electronic Measurement Technology, 2024, 47(9): 120-128.
[17] 刘敬敬, 孙凤乾, 张皓翔, 等. 基于YOLOv8的轻量化煤屑颗粒群实例分割方法[J]. 煤炭学报, 2024, 49(S2): 1310-1321.
LIU J J, SUN F Q, ZHANG H X, et al. Case segmentation method of lightweight coal particle swarm based on YOLOv8[J]. Journal of China Coal Society, 2024, 49(S2): 1310-1321.
[18] 贾岚, 徐雪环, 罗德弢, 等. 基于CrossStage-YOLO的跨阶段作物害虫检测[J]. 南京农业大学学报, 2025, 48(3): 744-754.
JIA L, XU X H, LUO D T, et al. Cross-stage crop pest dete-ction based on CrossStage-YOLO[J]. Journal of Nanjing Agricultural University, 2025, 48(3): 744-754.
[19] 田子建, 阳康, 吴佳奇, 等. 基于LMIENet图像增强的矿井下低光环境目标检测方法[J]. 煤炭科学技术, 2024, 52(5): 222-235.
TIAN Z J, YANG K, WU J Q, et al. LMIENet enhanced object detection method for low light environment in underground mines[J]. Coal Science and Technology, 2024, 52(5): 222-235.
[20] 丁玲, 缪小然, 胡建峰, 等. 改进YOLOv8s与DeepSORT的矿工帽带检测及人员跟踪[J]. 计算机工程与应用, 2024, 60(5): 328-335.
DING L, MIAO X R, HU J F, et al. Improved miner chin strap detection and personnel tracking with YOLOv8s and DeepSORT[J]. Computer Engineering and Applications, 2024, 60(5): 328-335.
[21] 韩康, 李敬兆, 陶荣颖. 基于改进YOLOv7和ByteTrack的煤矿关键岗位人员不安全行为识别[J]. 工矿自动化, 2024, 50(3): 82-91.
HAN K, LI J Z, TAO R Y. Recognition of unsafe behaviors of key position personnel in coal mines based on improved YOLOv7 and ByteTrack[J]. Journal of Mine Automation, 2024, 50(3): 82-91.
[22] CHEN J R, KAO S H, HE H, et al. Run, don’t walk: chasing higher FLOPS for faster neural networks[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2023: 12021-12031.
[23] OUYANG D, HE S, ZHAN J, et al. Efficient multi-scale attention module with cross-spatial learning[J]. arXiv:2305. 13563, 2023.
[24] MISRA D, NALAMADA T, ARASANIPALAI A U, et al. Rotate to attend: convolutional triplet attention module[C]//Proceedings of the IEEE Winter Conference on Applications of Computer Vision. Piscataway: IEEE, 2021: 3138-3147.
[25] ZHENG Z H, WANG P, LIU W, et al. Distance-IoU loss: faster and better learning for bounding box regression[C]// Proceedings of the AAAI Conference on Artificial Intelligence, 2020: 12993-13000.
[26] LIU C, WANG K, LI Q, et al. Powerful-IoU: more straightforward and faster bounding box regression loss with a nonmonotonic focusing mechanism[J]. Neural Networks, 2024, 170: 276-284.
[27] LIAO M H, WAN Z Y, YAO C, et al. Real-time scene text detection with differentiable binarization[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2020: 11474-11481.
[28] HOWARD A, SANDLER M, CHEN B, et al. Searching for MobileNetV3[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2019: 1314-1324.
[29] DU Y K, CHEN Z N, JIA C Y, et al. SVTR: scene text recognition with a single visual model[J]. arXiv:2205.00159, 2022.
[30] SHORTEN C, KHOSHGOFTAAR T M. A survey on image data augmentation for deep learning[J]. Journal of Big Data, 2019, 6(1): 60.
[31] WANG C Y, YEH I H, MARK LIAO H Y. YOLOv9: learning what you want to learn using programmable gradient information[C]//Proceedings of the European Conference on Computer Vision. Cham: Springer International Publishing, 2025: 1-21.
[32] WANG A , CHEN H , LIU L, et al. YOLOv10: real-time end-to-end object detection[J]. arXiv:2405.14458, 2024.
[33] TIAN Y, YE Q, DOERMANN D. YOLOv12: attention-centric real-time object detectors[J]. arXiv:2502.12524, 2025.