《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (7): 2210-2218.DOI: 10.11772/j.issn.1001-9081.2021040648

• 多媒体计算与计算机仿真 • 上一篇    

各向异性非极大值抑制在工业目标检测中的应用

张诗文1,2,3, 邓春华1,2,3(), 张俊雯1,2,3   

  1. 1.武汉科技大学 计算机科学与技术学院, 武汉 430065
    2.武汉科技大学 大数据科学与工程研究院, 武汉 430065
    3.智能信息处理与实时工业系统湖北省重点实验室(武汉科技大学), 武汉 430065
  • 收稿日期:2021-04-25 修回日期:2021-06-25 接受日期:2021-07-09 发布日期:2022-07-15 出版日期:2022-07-10
  • 通讯作者: 邓春华
  • 作者简介:张诗文(1997—),男,湖北建始人,硕士研究生,主要研究方向:计算机视觉、机器学习
    张俊雯(1997—),女,湖北荆门人,硕士研究生,主要研究方向:计算机视觉、机器学习。
  • 基金资助:
    国家自然科学基金资助项目(61806150)

Application of anisotropic non-maximum suppression in industrial target detection

Shiwen ZHANG1,2,3, Chunhua DENG1,2,3(), Junwen ZHANG1,2,3   

  1. 1.School of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan Hubei 430065,China
    2.Institute of Big Data Science and Engineering,Wuhan University of Science and Technology,Wuhan Hubei 430065,China
    3.Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System (Wuhan University of Science and Technology),Wuhan Hubei 430065,China
  • Received:2021-04-25 Revised:2021-06-25 Accepted:2021-07-09 Online:2022-07-15 Published:2022-07-10
  • Contact: Chunhua DENG
  • About author:ZHANG Shiwen, born in 1997, M. S. candidate. His research interests include computer vision, machine learning.
    ZHANG Junwen, born in 1997, M. S. candidate. Her research interests include computer vision, machine learning.
  • Supported by:
    National Natural Science Foundation of China(61806150)

摘要:

在某些固定的工业应用场景中,对目标检测算法的漏检容忍性非常低。然而,提升召回率的同时,目标周围容易规律性地产生一些无重叠的虚景框。传统的非极大值抑制(NMS)策略主要作用是抑制同一目标的多个重复检测框,无法解决上述问题。为此设计了一种各向异性NMS方法来对目标周围不同方向采取不同的抑制策略,从而有效消除规律性的虚景框。固定的工业场景中的目标形状和规律的虚景框往往具有一定关联性。为了促进各向异性NMS在不同方向的精确执行,设计了一种比例交并比(IoU)损失函数用来引导模型拟合目标的形状。此外,针对规则目标使用了一种自动标注的数据集增广方法,在降低人工标注工作量的同时扩大了数据集规模。实验结果表明,所提方法在轧辊凹槽检测数据集上的效果显著,应用于YOLO系列算法时在不降低速度的同时提升了检测精度。目前该算法已成功应用于某冷轧厂轧辊自动抓取的生产线。

关键词: 各向异性, 非极大值抑制, 交并比, 目标检测, YOLO

Abstract:

In certain fixed industrial application scenarios, the tolerance of the target detection algorithms to miss detection is very low. However, while increasing the recall, some non-overlapping virtual frames are likely to be regularly generated around the target. The traditional Non-Maximum Suppression (NMS) strategy has the main function to suppress multiple repeated detection frames of the same target, and cannot solve the above problem. To this end, an anisotropic NMS method was designed by adopting different suppression strategies for different directions around the target, and was able to effectively eliminate the regular virtual frames. The target shape and the regular virtual frame in a fixed industrial scene often have a certain relevance. In order to promote the accurate execution of anisotropic NMS in different directions, a ratio Intersection over Union (IoU) loss function was designed to guide the model to fit the shape of the target. In addition, an automatic labeling dataset augmentation method was used for the regular target, which reduced the workload of manual labeling and enlarged the scale of the dataset. Experimental results show that the proposed method has significant effects on the roll groove detection dataset, and when it is applied to the YOLO (You Only Look Once) series of algorithms, the detection precision is improved without reducing the speed. At present, the algorithm has been successfully applied to the production line of a cold rolling mill that automatically grabs rolls.

Key words: anisotropic, Non-Maximum Suppression (NMS), Intersection over Union (IoU), target detection, YOLO (You Only Look Once)

中图分类号: