检索结果-内蒙古大学图书馆

Review on applications of computer vision techniques for pipeline inspection

MEASUREMENT 2025年 252卷

作者： Liu, Yiming Moghaddas, Seyed A. Shi, Shuomang Huang, Ying Kong, Jun Bao, Yi Stevens Inst Technol Dept Civil Environm & Ocean Engn Hoboken NJ 07030 USA North Dakota State Univ Dept Civil Construct & Environm Engn Fargo ND 58102 USA North Dakota State Univ Dept Comp Sci Fargo ND 58102 USA

Pipeline inspection is critical for ensuring the safety, integrity, and efficiency of energy and water transportation systems. Anomalies such as leaks, corrosion, and structural deformations can result in severe issues for the environment, economy, and public safety. Traditional inspection methods, such as manual visual inspection, magnetic flux leakage, and ultrasonic testing, face challenges such as high cost, labor intensity, and limited automation. Computer vision has emerged as a promising solution to overcome these limitations by enabling automated, efficient, and cost-effective anomaly detection in pipelines. This paper systematically reviews the application of computer vision techniques for pipeline inspection, including data collection, image processing, deep learning-based analysis, digital twin-based management. Key influencing factors, challenges, and opportunities are discussed. The findings reveal that, with appropriate preprocessing, deep learning models achieved high accuracy in detecting cracks, corrosion, deformation, and dents, but their generalization performance was dependent on dataset quality, quantity, and preprocessing techniques. This review highlights the potential and barriers of computer vision to revolutionize pipeline inspection, providing a structured foundation for future research and applications.

关键词： Computer vision Deep learning Digital twin image processing Nondestructive evaluation (NDE) Pipeline inspection

来源：评论

学校读者我要写书评

暂无评论

Enhancing manufacturing process accuracy: A multidisciplinary approach integrating computer vision, machine learning, and control systems

引用

JOURNAL OF MANUFACTURING PROCESSES 2025年 142卷 453-467页

作者： Ramesh, Kaki Deshmukh, Sandip Ray, Tathagata Parimi, Chandu Birla Inst Sci & Technol Pilani Dept Mech Engn Hyderabad Campus Hyderabad 500078 Telangana India Birla Inst Sci & Technol Pilani Dept Comp Sci & Informat Syst Hyderabad Campus Hyderabad 500078 Telangana India Birla Inst Sci & Technol Pilani Dept Civil Engn Hyderabad Campus Hyderabad 500078 Telangana India

Manufacturing industries face significant challenges in producing high-quality, faultless products within limited timeframes. Conventional human-based inspection methods are still prone to errors and cannot guarantee precise component placement, potentially leading to product failures, user hazards, and substantial financial and reputational losses. This research presents a workflow to automate an inspection system that integrates computer vision, machine learning, image processing, and control systems to address these challenges. The proposed system employs a microcontroller and stepper motors to control a highly calibrated camera, enabling precise and efficient product inspection. At its core, the system utilizes the YOLOv5 model for object detection, specifically identifying hole marks and holes on products pre-assembly. This deep learning model was chosen for its real-time detection capabilities and high accuracy, achieving a mean Average Precision (mAP) of 0.95, which surpasses many current industry standards. Following object detection, advanced image processing techniques are applied to determine the precise position of detected features. Our approach achieves a notable error rate of 0.2 %, offering improvements over traditional inspection methods. Our system offers the potential to reduce inspection processing time and improve fault identification accuracy in real-time applications. Our research contributes to the field of industrial automation by introducing a seamless integration of state-of-the-art computer vision techniques with practical control systems. The system's modular design allows for easy adaptation to various manufacturing environments, benefiting industries with complex assembly processes, such as electronics, automotive manufacturing, etc. While the current implementation focuses on hole detection, future work will explore expanding the system's capabilities to identify a broader range of defects and adapt to different product types. This re

关键词： Deep learning image processing Assembly Manufacturing industries Quality assurance

来源：评论

学校读者我要写书评

暂无评论

An unsupervised learning method based on U-Net plus plus for low-light image enhancement

引用

SIGNAL image AND VIDEO processing 2025年第4期19卷 1-15页

作者： Wang, Xinghao Wang, Yu Zhou, Jian Liu, Jiaqi Gao, Yifan Wang, Yang Zheng, Jianbin Wuhan Univ Technol Sch Informat Engn Wuhan Peoples R China Wuhan Univ Technol Hubei Key Lab Broadband Wireless Commun & Sensor N Wuhan Peoples R China

In industrial settings like process monitoring and production line optimization, high-quality images play a critical role in managing operational details, rectifying procedural deviations, and improving strategic production outcomes. images captured in low-light conditions frequently exhibit color distortion, low contrast, and loss of detail, adversely affecting the accuracy and reliability of experimental data. To address these challenges, the study introduces an unsupervised learning method that utilizes the advanced U-Net + + architecture for enhancing low-light images. The advanced U-Net + + architecture, which replaces the traditional U-Net model, enhances deep feature extraction and learning mechanisms, effectively improving image processing for unaligned data. This approach substantially improves low-light photos in brightness, contrast, and color. Enhancements in image quality provide a solid foundation for more reliable and accurate analytical assessments in various applications. Experimental evidence shows that this method not only outperforms existing techniques in visual quality but also excels in key objective metrics, including Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), contrast difference, and image entropy ratio. This research significantly enhances image quality in low-light conditions using the U-Net + + architecture, improving both visual clarity and informational content crucial for accurate industrial experiments. The advancement promises to boost efficiency in industrial automation, quality inspection, and production management, thereby advancing technological development.

关键词： Low-light image enhancement Adversarial generative networks Unpaired data U-Net plus plus

来源：评论

学校读者我要写书评

暂无评论

Using Phased Array Ultrasound to Localize Probes During the inspection of Welds

IEEE OPEN JOURNAL OF INSTRUMENTATION AND MEASUREMENT

引用

IEEE OPEN JOURNAL OF INSTRUMENTATION AND MEASUREMENT 2023年 2卷 1-10页

作者： Gilmour, Adam Ulrichsen, Alexander Jackson, William Tabatabaeipour, Morteza Dobie, Gordon Macleod, Charles N. Murray, Paul Karkera, Benjamin Univ Strathclyde Dept Elect & Elect Engn Glasgow G1 1XW Scotland BAE Syst Maritime LA14 1AF Barrow In Fitness AK USA

In this article, an image processing-based localization system is developed for remote nondestructive evaluation of welds within industrial assets. Manual ultrasonic inspection of large-scale structures is often repetitive, time-consuming, and benefits greatly from robotic support, however, these robotic systems are often fixed to a single purpose, lack self-awareness of their surrounding environment, and can be limited to simple geometry. For the inspection of welds, which are often carried out using phased array ultrasonic testing, there is a reliance on the use of surface features for automated tracking such as the laser profiling of a weld cap. For the inspection of more complex geometry such as nonlinear or saddle welds, a more positionally sensitive method is required. The proposed system utilizes information already available to a nondestructive inspector in the form of live phased array ultrasonic images to estimate the location of the weld using nonsurface, volumetric data. Data is captured using a 64-element, 10-MHz phased array probe mounted to the end effector of a small robotic manipulator which increases the scope of applications due to its heightened flexibility when compared to on-the-market alternatives. Morphological operations are applied to the ultrasonic data to reduce the noise apparent from regions of parent material and promote the data reflected from grain boundaries within the weld material. Through a series of image processing techniques, it is possible to predict the position of a weld under inspection with an absolute mean positional error of $\mathrm {0.8 \text {m} \text { m} }$ . From this study, the localization system is to be embedded within a remote system for extensive data acquisition of welds on large structures.

关键词： Welding inspection Acoustics Probes Phased arrays Location awareness Steel automation image processing NDT&E phased array ultrasonic testing (PAUT) robotic inspection weld inspection weld localization

来源：评论

学校读者我要写书评

暂无评论

Survey on image-Oriented Transmission Line inspection via UAV

Survey on Image-Oriented Transmission Line Inspection via UA...

引用

2023 International Conference on Computer Science and automation Technology, CSAT 2023

作者： Quan, Ran Sun, Feng Hu, Xiaoping Yu, Miao Wang, Jinxu Li, Zixu State Grid Jilin Electric Power Co. Ltd. Changchun China Shenyang Jiuzhou Feixiang Tech. Co. Ltd Shenyang China Ccst Jilin University Changchun China

ISBN: (纸本)9798350371437

Unmanned aerial vehicle (UAV) is exploited gi-gantically to substitute the human beings in the wild natural working environments to face the challenges of severe climate situations and low job efficiency. This paper surveys the reported applications of UAV in long-distance transmission line inspection via the research and techniques in image processing. It discusses the major pros and cons of the state of the art methods and discloses the future working directions to benefit the research domain. © 2023 IEEE.

关键词： Electric lines

来源：评论

学校读者我要写书评

暂无评论

Viewpoint Planning for Fuel Leak Detection in Rail-Mounted inspection Robot 5

Viewpoint Planning for Fuel Leak Detection in Rail-Mounted I...

引用

5th International Conference on Machine Learning and Computer Application, ICMLCA 2024

作者： Zhu, Sheng Zhang, Shuyue Zhang, Kanjian Wei, Haikun School of Automation Southeast University Nanjing China

ISBN: (纸本)9798331530334

This paper proposes a fuel leak detection method designed for unmanned Antarctic energy platforms, applied to a rail-mounted inspection robot. By introducing a viewpoint planning strategy and utilizing vector geometry to calculate perspective features, the robot can automatically optimize its camera angles to capture images from the best viewpoints. Additionally, this paper employs multi-scale Retinex and high-pass filtering techniques to enhance image processing and achieve uniform lighting, significantly improving the accuracy of fuel leak detection. Experimental results demonstrate that this method exhibits excellent detection performance in simulated environments, validating its feasibility for practical applications. © 2024 IEEE.

关键词： Robot programming

来源：评论

学校读者我要写书评

暂无评论

A Review on image processing Tecniques for In-Pipe Robots 5th

A Review on Image Processing Tecniques for In-Pipe Robots

引用

5th International Conference on Recent Trends in Machine Learning, IoT, Smart Cities, and applications, ICMISC 2024

作者： Suryawanshi, Shubhangi Chavan, Amol Bhosale, Digvijay G. Mahakal, Shivam Jakate, Govind Kanojiya, Abhay Department of Artificial Intelligence and Data Science Dr. D. Y. Patil Institute of Technology Maharashtra Pune411 018 India Department of Mechatronics Engineering Rajarambapu Institute of Technology Shivaji University RajaramnagarKolhapur Maharashtra 415 414 India Department of Mechanical Engineering Dr. D. Y. Patil Institute of Technology Maharashtra Pune411 018 India Department of Automation and Robotics Dr. D. Y Patil Institute of Technology Maharashtra Pune411 018 India

ISBN: (纸本)9789819788644

Pipelines are vital infrastructure for the transportation of water, sewage, oil, and gas because they offer a better safety option and fit into the minimum environmental hazardous than other modes. But pipelines are aging and getting damaged all the time, which can lead to significant loss of resources and contamination of the environment. Regular maintenance and inspections are necessary for pipelines to operate efficiently and prevent overspending. This paper reviews the core techniques used in each of the four missions—localization, mapping, navigation, and inspection—that these robots have completed. Additionally, because image processing has become a widely used and crucial strategy for completing all of the aforementioned missions, we made the decision to review the thorough classification of techniques for image processing in a separate section. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Navigation

来源：评论

学校读者我要写书评

暂无评论

Enhancing Agricultural Productivity: A Sensor Fusion, Spectroscopic Near Infrared, and AI-Based Framework for Monitoring Fruit Ripeness 21

Enhancing Agricultural Productivity: A Sensor Fusion, Spectr...

引用

21st International Bhurban Conference on Applied Sciences and Technology, IBCAST 2024

作者： Ambreen, Hira Imran, Muhammad Diyin, Tang Zhang, Chongsheng Sir Syed CASE Institute of Technology Islamabad Pakistan School of Automation Science and Electrical Engineering Beihang University Aeronautics and Astronautics China Henan Univerity Department of EE Kaifeng475001 China School of Automation Science and Electrical Engineering Beihang University Aeronautics and Astronautics Department of Detection and Automation Engineering China School of Computer & Information Engineering Henan University Kaifeng475001 China

ISBN: (纸本)9798331516680

Fruit ripeness evaluation is essential for harvest optimization and post-harvest loss prevention. Traditional visual inspection is inefficient. This study employs sensor fusion and AI algorithms to evaluate fruit maturity in real-time using gas sensors (i.e., MQ-3, MQ-6, MQ-8, MQ-135) and Near-Infrared spectral sensors (i.e., GY-AS7263). Primarily, data pro-cessing-sensor fusion, data preparation, and AI model application-is crucial to the proposed framework. Subsequently, sensor fusion approaches improve data accuracy by combining signals from various sensors. Secondly, data pretreatment techniques including noise reduction, normalization, and feature extraction prepare data for analysis. Thirdly, we use powerful machine learning algorithms such as Random Forest to judge maturity more accurately and reliably than conventional techniques. Lastly, integrating NIR spectroscopy with gas sensor data resolves cost and model transferability issues, enhancing accuracy and acces-sibility for agricultural use. Static image processing techniques provide challenges for real-time applications, including expensive computer costs and lower reliability. The proposed framework based on multi-sensor fusion system enhances agricultural applications' versatility, scalability, and reactivity, surpassing existing methodologies. © 2024 IEEE.

关键词： Fruits

来源：评论

学校读者我要写书评

暂无评论

PCB Component inspection System Using Raspberry Pi

PCB Component Inspection System Using Raspberry Pi

引用

2023 IEEE Technology and Engineering Management Conference - Asia Pacific, TEMSCON-ASPAC 2023

作者： Deepa, M. Kaniskaa, Ms Hariprasath, K. Eashwar, S. Subbu, M. Psg Institute of Technology and Applied Research Department of Electronics and Communication Engineering Coimbatore India

ISBN: (纸本)9798350384659

High-performance vision inspection systems that can analyze enormous volumes of unstructured data are in high demand in the field of industrial automation. This project outlines a novel strategy to meet this requirement by emphasizing the application side of image collecting and processing rather than the underlying *** is built to function flawlessly in industrial environments while collecting photos from various sources. These photos go through a number of sophisticated image processing techniques because they are frequently filled with complex and unstructured data. The system's strengths cover a wide range of industrial fields, such as dimensions analysis, defect identification, and object recognition. It offers quick data processing and high levels of accuracy through careful algorithm design and optimization, making it a vital asset in improving industrial *** industrial applications, this system handles unorganised data exceptionally well. Industries are able to improve product quality, lower faults, and streamline operations because to its powerful picture analysis capabilities. We highlight the system's potential to lead to substantial developments and enhancements across many industrial sectors by concentrating on its actual applications. The proposed system uses Raspberry Pi 2 board as the processor along with the image processing algorithms and R-CNN algorithm using OpenCV library in Python for comparison of the images. Thus, the model is trained effectively to meet the requirements and inspect PCB components successfully. © 2023 IEEE.

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

A Novel image Stitching Method Based on an Unsupervised Deep Learning Algorithm Considering Homography Estimation and Networks

A Novel Image Stitching Method Based on an Unsupervised Deep...

引用

2024 International Conference on Remote Sensing, Mapping, and image processing, RSMIP 2024

作者： Hu, Hao Sun, Ting Yu, Shijie Deng, Shuixin Chen, Baohua School of Instrument Science and Opto-Electrics Engineering Beijing Information Science and Technology University Beijing100192 China Laboratory of Intelligent Microsystems Beijing Information Science and technology University Beijing100192 China Department of automation Tsinghua University Beijing100084 China

ISBN: (纸本)9781510680012

With the continuous development of digital photography technology and quality inspection, the demand for image stitching in practical applications is increasing. Traditional image stitching algorithms employ a variety of hand-designed methods for feature extraction, matching, and optimization. However, these traditional feature-based image stitching techniques heavily rely on feature extraction and may not perform well in scenarios with limited features. Current image stitching solutions based on supervised deep learning lack relevant data sets, and labeling data is relatively cumbersome, making supervised deep learning methods unreliable. At the same time, the rise of unsupervised deep learning algorithms provides new ideas for image stitching. We use unsupervised homography estimation to provide information about the geometric relationship between images, Stitching-Domain Transformer Layer to align feature maps, warp and generate masks, it helps to enhance the reality and continuity of splicing. We simultaneously utilize a pre-trained deep learning model (VGG) for feature extraction. We adjust the smoothness loss term to ensure smoother transitions within the stitching areas. Throughout the training process, we continuously optimize the number of convolutional layers, channels, and network depth to achieve optimal results. The superiority of the unsupervised learning algorithm compared to other classic algorithms was verified through experiments. Finally, we discussed the challenges and future applications of unsupervised deep learning in image stitching. © 2024 SPIE.

关键词： Feature extraction

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：