王 宸,杨 帅,周 林,华珀玺,王生怀,吕 江.基于自适应多尺度特征融合网络的金属齿轮端面缺陷检测方法研究[J].电子测量与仪器学报,2023,37(10):153-163 |
基于自适应多尺度特征融合网络的金属齿轮端面缺陷检测方法研究 |
Research on metal gear end-face defect detection method based on adaptive multi-scale feature fusion network |
|
DOI: |
中文关键词: 齿轮端面 YOLOv5s CBAM BiFPN 缺陷检测 |
英文关键词:gear end-face YOLOv5s CBAM BiFPN defect detection |
基金项目:国家自然科学基金(51475150)、教育部人文社科项目(20YJCZH150)、湖北省重点研发计划项目(2021BAA056) 、湖北省高等学校中青年科技创新团队计划项目(T20200018)、湖北省社科基金(21Q174)、湖北汽车工业学院博士基金(BK201905)项目资助 |
|
|
摘要点击次数: 749 |
全文下载次数: 715 |
中文摘要: |
针对目前金属齿轮端面结构复杂,导致缺陷的小目标占比度高和尺度变化大引起的检测准确度低,难以满足企业实时
在线检测需求等问题。 本文基于 YOLOv5s 网络提出了一种基于自适应多尺度特征融合网络的金属齿轮端面缺陷检测方法
(YOLO-Gear)。 首先,搭建了一个齿轮端面缺陷检测试验台,并制作了齿轮端面缺陷数据集。 然后,提出了自适应卷积注意力
模块(convolutional block attention module-C3,CBAM-C3),CBAM-C3 通过将通道注意力(channel attention module,CAM)和空间注
意力(spartial attention module,SAM)相结合加强了对金属齿轮缺陷小目标缺陷自适应的特征学习与特征提取,及时对模型中的
权重参数进行学习和优化,提高了模型对小目标缺陷的检测准确度;最后,提出了重复加权双向特征金字塔网络( bidirectional
feature pyramid network,BiFPN),通过自适应控制不同尺度的特征图之间的融合程度,提高了模型对缺陷多尺度检测能力。 试
验表明,YOLO-Gear 模型在齿轮端面缺陷测试集上的平均精度达到了 99. 2%,F1 值为 0. 99,FPS 值为 33。 相较于其他深度学习
模型,本文提出的 YOLO-Gear 模型提高了检测的精度和效率,能够满足企业的实时在线检测需求。 |
英文摘要: |
The high proportion and large-scale variation of small targets with defects caused by the complex structure of metal gear end
faces have led to low detection accuracy, making it difficult to meet the real-time online detection needs of enterprises. In this paper, we
propose a metal gear end face defect detection method based on an adaptive multi-scale feature fusion network (YOLO-Gear) using the
YOLOv5s network. Firstly, we establish a gear end face defect detection test platform and create a gear end face defect dataset. Then,
we introduce the adaptive convolutional block attention module (CBAM-C3) which combines channel attention module ( CAM) and
spatial attention module (SAM) to enhance the adaptive feature learning and extraction for small target defects in metal gears, effectively
improving the detection accuracy of the model for small target defects. Finally, we propose the bidirectional feature pyramid network
(BiFPN), which repetitively weights and fuses features from different scales, thereby improving the model’s ability to detect defects at
multiple scales. Experimental results demonstrate that the YOLO-Gear model achieves an average precision of 99. 2%, an F1 score of
0. 99, and an FPS value of 33 on the gear end face defect test set. Compared to other deep learning models, the proposed YOLO-Gear
model in this paper improves both detection accuracy and efficiency, meeting the real-time online detection needs of enterprises. |
查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|