基于融合残差门控重建的轴承缺陷检测算法研究OA
Research on Bearing Defect Detection Algorithm Based on Fused Residual Gated Reconstruction
针对轴承表面缺陷检测中微小特征易被背景淹没、复杂形态定位不准,提出 YOLOv12改进算法.引入残差门控空间-通道重建模块,自适应调节融合强度,增强微小缺陷感知;加权组合SIoU与CIoU作为回归损失,协同优化角度与形状,兼顾方向敏感与不规则定位.所提模型以2 546 035参数量和6.1 GFLOPs计算量,实现了88.6%的 mAP50和64.2%的 mAP95,相比基准YOLOv12n分别提升2.6百分点和2.1百分点.在相近或更小的模型规模下,所提方法取得了优于 YOLOv8s、YOLOv11s、YOLOv12s等small版本以及RT-DETR、Deformable-DETR的检测精度,同时保持了极高的计算效率.
To address the issues of small defect features being easily overwhelmed by background clut-ter and inaccurate localization of complex morphologies in bearing surface defect detection,an improved YOLOv12 algorithm is proposed.A residual-gated spatial-channel reconstruction module is introduced to adaptively adjust the fusion intensity,thereby enhancing the perception of small defects.The regression loss is formulated as a weighted combination of SIoU and CIoU,which synergistically optimizes angular and shape attributes while balancing directional sensitivity and irregular localization.The proposed model achieves 88.6%mAP50 and 64.2%mAP95 with only 2 546 035 parameters and 6.1 GFLOPs,outperfor-ming the baseline YOLOv12n by 2.6 percentage points and 2.1 percentage points,respectively.With a comparable or even smaller model scale,the proposed method obtains higher detection accuracy than the small versions of YOLOv8s,YOLOv11s,and YOLOv12s,as well as RT-DETR and Deformable-DETR,while maintaining exceptionally high computational efficiency.
王勇;刘献泓;刘盛;周丹;汤志伟;周锋
江苏省特种设备安全监督检验研究院盐城分院,江苏 盐城 224051盐城工学院信息工程学院,江苏 盐城 224051江苏省特种设备安全监督检验研究院盐城分院,江苏 盐城 224051江苏省特种设备安全监督检验研究院盐城分院,江苏 盐城 224051盐城工学院信息工程学院,江苏 盐城 224051盐城工学院信息工程学院,江苏 盐城 224051||盐城工学院光电信息技术研究所,江苏 盐城 224051
机械制造
轴承表面缺陷检测YOLOv12残差门控空间-通道重建模块SIoU-CIoU组合损失
bearing surface defect detectionYOLOv12residual-gated spatial-channel reconstruction moduleSIoU-CIoU combined loss
《机械与电子》 2026 (7)
53-60,70,9
江苏省特检院资助项目(KJ(Y)202633)盐城市重点研发计划工业领域竞争资助项目(YCBG2024023)盐城市应用基础研究计划资助项目(YCBK2025007)江苏省研究生科研与实践创新计划资助项目(SJCX24_2153,SJCX25_2202)
评论