Knowledge Commons of Institute of Automation,CAS
Industrial WeakScratches Inspection Based on Multi-Feature Fusion Network | |
Tao Xian; Zhang DP(张大朋); Hou wei; Ma wenzhi; Xu De | |
发表期刊 | IEEE Transaction on Instrumentation and Measurement |
2020 | |
期号 | 1页码:1-14 |
摘要 | Scratches are one of the most common defects in industrial manufacturing. Weak scratches in the industrial environment have an ambiguous edge, low contrast, large span, and unfixed shape, which brings difficulty for automatic defect detection. Recently, many existing visual inspection methods based on deep learning cannot completely and effectively inspect industrial weak scratches due to the lack of discriminative features and sufficient spatial detail. In this paper, a novel DeepScratchNet is proposed for automatic weak scratch detection by aggregating rich multi-dimensional feature for scratch representation. To obtain rich features, a pre-trained ResNet block as a feature extractor is proposed in this paper. In order to highlight features of scratch and weaken the noise, an attention feature fusion block (AFB) is proposed, which densely fuses high-level semantic features with low-level details features using dual-attention mechanism. Due to the long span and connectivity of the weak scratches, a context fusion block (CFB) is proposedto learn the complete context.To further improve the scratch segmentation performance, the auxiliary loss is integrated into the proposed network. The proposed DeepScratchNet outperforms traditional and other state-of-the-art deep learning-based methods on three given real-world industrial datasets with mIoU over 0.8005, 0.812 and 0.9286. The experimental results demonstrate that DeepScratchNetachievesgoodgeneralizationcapabilities |
关键词 | Weak scratch inspection, Defect Detection, MultipleFeatureFusion,DeepLearning,MachineVision |
七大方向——子方向分类 | 目标检测、跟踪与识别 |
文献类型 | 期刊论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/40630 |
专题 | 中科院工业视觉智能装备工程实验室_精密感知与控制 |
通讯作者 | Tao Xian |
作者单位 | Institute of Automation, Chinese Academy of Science |
推荐引用方式 GB/T 7714 | Tao Xian,Zhang DP,Hou wei,et al. Industrial WeakScratches Inspection Based on Multi-Feature Fusion Network[J]. IEEE Transaction on Instrumentation and Measurement,2020(1):1-14. |
APA | Tao Xian,Zhang DP,Hou wei,Ma wenzhi,&Xu De.(2020).Industrial WeakScratches Inspection Based on Multi-Feature Fusion Network.IEEE Transaction on Instrumentation and Measurement(1),1-14. |
MLA | Tao Xian,et al."Industrial WeakScratches Inspection Based on Multi-Feature Fusion Network".IEEE Transaction on Instrumentation and Measurement .1(2020):1-14. |
条目包含的文件 | ||||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
Industrial Weak Scra(5789KB) | 期刊论文 | 作者接受稿 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论