检索结果-内蒙古大学图书馆

M-KANUNet: enhanced defect segmentation in X-ray images of copper pipe welds via multi-scale representation and Kolmogorov-Arnold Networks

引用

VISUAL COMPUTER 2025年 1-14页

作者： Yang, Lihua Zhao, Jinxian Wang, Ziming Liu, Yuheng Chi, Dazhao Harbin Inst Technol State Key Lab Precis Welding & Joining Mat & Struc Harbin 150001 Peoples R China Heilongjiang Univ Sci & Technol Sch Comp & Informat Engn Harbin 150022 Peoples R China

In the X-ray digital images of small-diameter copper tube soldering structures, the significant variations in defect size, morphology, and contrast make automatic defect recognition challenging. To address this problem, this study proposes M-KANUNet, an improved TransUNet-based defect segmentation model, designed for segmenting lack of penetration defects in X-ray images of copper tube welds. First, the ResNet module in the encoder part of the baseline TransUNet model is replaced with a Res2Net module combined with the SimAM attention mechanism, enhancing the model's ability to extract multi-scale features and capture local defect details in complex backgrounds. Second, leveraging the strengths of the Kolmogorov-Arnold Networks (KANs), the KAN-Transformer module is proposed to improve the model's ability to process complex data by using learnable activation functions at the network's edges. Experimental results on a constructed dataset of 496 X-ray images show that, compared to the baseline TransUNet model, M-KANUNet achieves improvements of 1.84% in intersection over union (IoU), 1.24% in recall (R), and 2.29% in precision (P), demonstrating higher reliability. Additionally, the frames per second(FPS) of M-KANUNet increases by 10 f/s, and the number of model parameters is reduced by 79.87M. The source code is available at https://***/Lihua-Y/M-KANUNet.

关键词： X-ray image defect segmentation Kolmogorov-Arnold Networks SimAM attention mechanism Res2Net TransUNet

来源：评论

学校读者我要写书评

暂无评论

Progressive Feature Enhancement Network for Surface defect segmentation

引用

IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE 2025年

作者： Yan, Feng Jiang, Xiaoheng Zhang, Yunxia Lu, Yang Nan, Xiaofei He, Shuo Xu, Mingliang Zhengzhou Univ Sch Comp Sci & Artificial Intelligence Zhengzhou 450001 Peoples R China Zhengzhou Univ Engn Res Ctr Intelligent Swarm Syst Minist Educ Zhengzhou 450001 Peoples R China Zhengzhou Univ Natl Supercomp Ctr Zhengzhou Zhengzhou 450001 Peoples R China

Surface defect detection is critical for maintaining the high quality of industrial products. However, defect detection is confronted with challenges, such as diverse defect types and scales, low contrast, and complex backgrounds. To tackle the problems, we propose a Progressive Feature Enhancement Network (PFENet), which aims to gradually strengthen the representation of features through semantic-guided Single-scale Feature Enhancement (SFE) module and Cross-scale Feature Enhancement (CFE) module. Specifically, SFE highlights defect semantic information of multi-level features by exploiting spatial similarities between the features and high-level features. CFE adaptively selects important defect information and suppresses redundant information through the mutual interaction of cross-level features. The mutual interaction enlarges the difference between foreground and background and facilitates learning more discriminative defect features for complex defects. Extensive experiments on three publicly available defect datasets, magnetic tile (MT), NEU-Seg, and Road defect dataset demonstrate that the proposed method achieves state-of-the-art performance.

关键词： Feature extraction Semantics defect detection Transformers Computational modeling Convolutional codes Surface morphology Shape Neural networks Interference Cross-scale feature fusion defect segmentation feature enhancement semantic guidance

来源：评论

学校读者我要写书评

暂无评论

LIEnet: Local information enhancement network for multi-scale defect segmentation of optical module base cavity

引用

MEASUREMENT 2025年 250卷

作者： Yang, Dezhi Liu, Jian Suo, Xinyu Chen, Ning Li, Rong Hunan Univ Coll Mech & Vehicle Engn Changsha 410082 Peoples R China

The optical module is the core component for photoelectric conversion in the optical fiber communication process. The defects of the optical module base cavity seriously affect the installation quality and service life of the optical module. However, current segmentation methods are difficult to apply to multi-scale defects of the optical module base cavity. Small defects are often easily ignored or difficult to accurately segment. To address this issue, this paper proposes a novel local information enhancement network (LIEnet). First, the MobileViTv3 model is introduced as the backbone to enhance the interaction of global context information. Then a local information enhancement structure is proposed to enhance the fine segmentation ability. Finally, a combined cross-entropy loss function is constructed to alleviate the imbalance between foreground and background class samples. In the experiments on the self-built dataset and two public defect datasets, the LIEnet has better segmentation performance than other semantic segmentation methods.

关键词： Optical module base defect segmentation Local information enhancement Machine vision

来源：评论

学校读者我要写书评

暂无评论

MIAMS: Weld defect segmentation in time-of-flight diffraction images based on multi-image attention and main structure analysis

引用

MEASUREMENT 2025年 246卷

作者： Yang, Deyan Jiang, Hongquan Yue, Kun Zhi, Zelin Jing, Deqiang Gao, Jianmin He, Ziyong Wang, Hui Xi An Jiao Tong Univ State Key Lab Mfg Syst Engn Xian 710049 Peoples R China Shaanxi Special Equipment Inspection & Testing Ins Xian 710048 Peoples R China Dongfang Elect Machinery Co Ltd Deyang 618000 Peoples R China

Weld defect segmentation is crucial for realizing the defect rating of welded products and for evaluating the welding quality. However, the existing methods cannot accurately segment weld defect regions in time-of-flight diffraction (TOFD) images owing to blurring, weak texture variation, and surrounding noise and interference fringes. To address these issues, we propose a weld defect segmentation method based on multi-image attention and main structure analysis (MIAMS) for TOFD images, which includes three modules: multi-image feature fusion, two-branch feature extraction, and segmentation mask prediction. Additionally, we perform experimental validation using the weld TOFD images of large-scale spherical pressure tanks. The results indicate that MIAMS outperforms the existing semantic segmentation methods with a mean intersection over union and mean F-score of 86.18% and 92.28%, respectively. MIAMS is lightweight, and the results of defect segmentation closely match the results of expert evaluation in non-destructive testing, improving the quantitative accuracy of the defects.

关键词： Time-of-flight diffraction defect segmentation Multi-image attention Main structural analysis

来源：评论

学校读者我要写书评

暂无评论

YOLOSeg with applications to wafer die particle defect segmentation

引用

SCIENTIFIC REPORTS 2025年第1期15卷 1-14页

作者： Li, Yen-Ting Chan, Yu-Cheng Huang, Chen-Che Hsu, Yu-Chang Chen, Ssu-Han Circle AI Inc Taipei 114 Taiwan Ming Chi Univ Technol Ctr Artificial Intelligence & Data Sci New Taipei City 243 Taiwan Cathay Financial Holdings Co Ltd Taipei 106 Taiwan Ming Chi Univ Technol Dept Ind Engn & Management New Taipei City 243 Taiwan

This study develops the you only look once segmentation (YOLOSeg), an end-to-end instance segmentation model, with applications to segment small particle defects embedded on a wafer die. YOLOSeg uses YOLOv5s as the basis and extends a UNet-like structure to form the segmentation head. YOLOSeg can predict not only bounding boxes of particle defects but also the corresponding bounding polygons. Furthermore, YOLOSeg also attempts to obtain a set of better weights by combining with several training tricks such as freezing layers, switching mask loss, using auto-anchor and introducing denoising diffusion probabilistic models (DDPM) image augmentation. The experiment results on the testing image set show that YOLOSeg's average precision (AP) and intersection over union (IoU) are as high as 0.821 and 0.732 respectively. Even when the sizes of particle defects are extremely small, the performance of YOLOSeg is far superior to current instance segmentation models such as mask R-CNN, YOLACT, YUSEG, and Ultralytics's YOLOv5s-segmentation. Additionally, preparing the training image set for YOLOSeg is time-saving because it needs neither to collect a large number of defective samples, nor to annotate pseudo defects, nor to design hand-craft features.

关键词： Auto-annotation defect segmentation Wafer die You only look once (YOLO) Denoising diffusion probabilistic models (DDPM)

来源：评论

学校读者我要写书评

暂无评论

Deep learning driven silicon wafer defect segmentation and classification

引用

METHODSX 2025年 14卷 103158页

作者： Ingle, Rohan Shahade, Aniket K. Gaikwad, Mayur Patil, Shruti Symbiosis Int Symbiosis Inst Technol Pune Maharashtra India

Integrated Circuits are made of various transistors that are embedded on a silicon wafer, these wafers are difficult to process and hence are prone to defects. defecting these defects manually is a time consuming and labour-intensive task and hence automation is necessary. Deep Learning approach is better suited in this case as it is able to generalize defects if trained properly and can be a solution to segmentation and classification of defects automatically. The segmentation model mentioned in this study achieved a Mean Absolute Error (MAE) of 0.0036, a Root Mean Squared Error (RMSE) of 0.0576, a Dice Index (DSC) of 0.7731, and an Intersection over Union (IoU) of 0.6590. The classification model achieved 0.9705 Accuracy, 0.9678 Precision, 0.9705 Recall, and 0.9676 F1 Score. In order to make this process a more interactive, an LLM with Q&A capabilities was integrated to solve any doubts and answer any questions regarding defects in wafers. This approach helps automate the detection process thus improving quality of end product. center dot Successful and precise defect segmentation and classification using Deep Learning was achieved. center dot High-intensity regions after post-processing. center dot An LLM offering defect analysis and guidance was streamlined.

关键词： Silicon wafers Integrated circuit defect segmentation Deep learning Wafer defects Quality management Image segmentation

来源：评论

学校读者我要写书评

暂无评论

defect focused Harris3D & boundary fine-tuning optimized region growing: Lithium battery pole piece defect segmentation

引用

MEASUREMENT 2025年 242卷

作者： Ma, Ruijie Li, Chen Xing, Yibo Wang, Siyao Ma, Rui Feng, Feng Qian, Xiang Wang, Xiaohao Li, Xinghui Tsinghua Univ Shenzhen Int Grad Sch Shenzhen 518055 Peoples R China Tsinghua Univ Tsinghua Berkeley Shenzhen Inst Shenzhen 518055 Peoples R China

Asa vital component of lithium battery, the pole piece is prone to defects during the production process, which adversely affects performance and even causes safety accidents. Therefore, conducting quality inspections for lithium battery pole pieces is crucial. In this research, an optimized region growing algorithm based on 3D point cloud data is proposed to automatically detect pole piece defects. Specifically, a potential defect- focused Harris3D method (DFHarris3D) for keypoint detection and a boundary fine-tuning algorithm (BFT) are introduced. Comparative experimental results prove that our method outperforms others in terms of completeness, accuracy, and robustness. Therefore, this novel non-destructive detection technique demonstrates the potential to accurately segment surface defects in lithium battery pole pieces, which can be integrated into smart factories for fully automated defect detection.

关键词： defect segmentation 3D point cloud Region growing Lithium battery

来源：评论

学校读者我要写书评

暂无评论

Cross-supervised contrastive learning domain adaptation network for steel defect segmentation

引用

ADVANCED ENGINEERING INFORMATICS 2025年 64卷

作者： Zhao, Lixiang Zhang, Yulong Duan, Jie Yu, Jianbo Tongji Univ Sch Mech Engn Shanghai 201804 Peoples R China

defect segmentation is a vital approach to ensure the quality of steel. Convolutional neural network (CNN) has significantly enhanced steel defect segmentation performance. However, in practical scenario, steel defects exhibit great difference in appearance and background due to complex working conditions. It is challenging to perform steel defect segmentation under polytopic conditions. In this article, a cross-supervised contrastive learning domain adaptation network (CSCLDAN) is proposed for steel defect segmentation. Firstly, to alleviate domain shift posed by variations in defect appearance, a cross-supervised learning network that leverages CNN and Transformer is proposed to extract comprehensive features by facilitating the interaction of local and global information from different domains simultaneously. Secondly, for the domain shift caused by contextual difference, self-aware consistency contrastive learning is proposed to maintain consistency of the same defect features under different contexts. Finally, bidirectional pixel feature alignment loss is proposed to achieve self- aware consistency, which enables features from positive samples closer and separates features from negative samples. The effectiveness of CSCLDAN is verified on four typical steel defect datasets and the testing results demonstrate that CSCLDAN outperforms other state-of-the-art domain adaptation models.

关键词： Steel defect defect segmentation Domain adaptation Deep learning Contrastive learning

来源：评论

学校读者我要写书评

暂无评论

Leveraging Vision Foundation Model via PConv-Based Fine-Tuning with Automated Prompter for defect segmentation

引用

SENSORS 2025年第8期25卷 2417-2417页

作者： Jiang, Yifan Chen, Jinshui Lu, Jiangang Zhejiang Univ Coll Control Sci & Engn State Key Lab Ind Control Technol Hangzhou 310027 Peoples R China

In industrial scenarios, image segmentation is essential for accurately identifying defect regions. Recently, the emergence of foundation models driven by powerful computational resources and large-scale training data has brought about a paradigm shift in deep learning-based image segmentation. The Segment Anything Model (SAM) has shown exceptional performance across various downstream tasks, owing to its vast semantic knowledge and strong generalization capabilities. However, the feature distribution discrepancy, reliance on manually labeled prompts, and limited category information of SAM reduce its scalability in industrial settings. To address these issues, we propose PA-SAM, an industrial defect segmentation framework based on SAM. Firstly, to bridge the gap between SAM's pre-training data and distinct characteristics of industrial defects, we introduce a parameter-efficient fine-tuning (PEFT) technique incorporating lightweight Multi-Scale Partial Convolution Aggregation (MSPCA) into Low-Rank Adaptation (LoRA), named MSPCA-LoRA, which effectively enhances the image encoder's sensitivity to prior knowledge biases, while maintaining PEFT efficiency. Furthermore, we present the Image-to-Prompt Embedding Generator (IPEG), which utilizes image embeddings to autonomously create high-quality prompt embeddings for directing mask segmentation, eliminating the limitations of manually provided prompts. Finally, we apply effective refinements to SAM's mask decoder, transforming SAM into an end-to-end semantic segmentation framework. On two real-world defect segmentation datasets, PA-SAM achieves mean Intersections over Union of 73.87% and 68.30%, as well as mean Dice coefficients of 84.90% and 80.22%, outperforming other state-of-the-art algorithms, further demonstrating its robust generalization and application potential.

关键词： defect segmentation vision foundation model segment anything model low-rank adaptation partial convolution automated prompter parameter-efficient fine-tuning

来源：评论

学校读者我要写书评

暂无评论

Detection Model for Automatic defect Quantification and segmentation for Stepped Eddy Current Thermography

引用

IEEE SENSORS JOURNAL 2025年第7期25卷 10784-10799页

作者： Gao, Yuan Zhang, Liang Liang, Zheng Zheng, Ting Deng, Xiong Chen, Xin Southwest Petr Univ Sch Mech Engn Chengdu 610500 Peoples R China Luzhou Vocat & Tech Coll Sch Intelligent Mfg & Automot Engn Luzhou 646000 Sichuan Peoples R China Southwest Petr Univ Petr Engn Sch Chengdu 610500 Sichuan Peoples R China

The stepped eddy current thermography (SECT) nondestructive testing (NDT) technique is characterized by long heating time and nonuniform temperature rise. This causes it to remain challenging to quantify the geometrical features of defects on the inner wall of the tank roof and the outer wall of the tank bottom in oil and gas storage tanks. This article proposes a combined model for compressing and reconstructing thermal image sequences: the skewness model combined with the improved Gaussian adaptive background estimation algorithm (SM-IGABEA) for quantifying the defect morphology. The combined model is coupled with the first-order differential max-min method to quantify the width and height of defects accurately. The combined model combined with the first-order differential mean method can accurately segment defects. A mathematical model for predicting the residual depth (RD) of steel plates is developed to describe the relationship between the geometric characteristics of defects and the mean value of skewness. Finally, the generalization of SM-IGABEA is verified by elliptical defects. The results show that various combinatorial models and quantization methods are proposed for the defect measurement task. The measurement accuracy and stability of SM-IGABEA significantly outperform the mainstream compressive reconstruction algorithms.

关键词： Steel Eddy currents Sensors Infrared imaging Heating systems Temperature sensors Time series analysis Surface morphology Reconstruction algorithms Oils defect quantification study defect segmentation nondestructive testing (NDT) skewness model stepped eddy current thermography (SECT)

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：