检索结果-内蒙古大学图书馆

Automated steel corrosion detection and damage assessment using yolov5 algorithm

SHIPS AND OFFSHORE STRUCTURES 2025年

作者： Mu, Jianwu Yang, Yang Chu, Jinze Lv, Wanyang Hu, Kaixuan Dalian Univ Technol Sch Naval Architecture & Ocean Engn Dalian Liaoning Peoples R China

Current ship corrosion inspections rely heavily on manual observation, which is subjective, inefficient, and difficult in inaccessible areas. This paper proposes a method for identifying steel corrosion and determining damage levels using the yolov5 (You Only Look Once, Version 5) algorithm and its enhancements. The method increases the steel corrosion dataset and applies yolo for corrosion image recognition. Converting images to greyscale reduces the impact of colour on corrosion detection, while Gaussian filtering eliminates noise. A segmented linear transformation enhances contrast between corroded and non-corroded areas. To detect corrosion, a Otsu's method is used, and expansion operations reduce noise in small corrosion regions. The 8-neighborhood method then counts and analyses corrosion areas, providing details on number, size, and location. By combining these techniques with the DOP (Degree of Pitting Corrosion) index and damage classification, the method quantitatively evaluates corrosion damage on ship surfaces. Significantly improving detection efficiency and accuracy.

关键词： Corrosion damage identification convolutional neural networks yolo algorithm feature extraction

来源：评论

学校读者我要写书评

暂无评论

Enhancing autonomous pavement crack detection: Optimizing yolov5s algorithm with advanced deep learning techniques

引用

MEASUREMENT 2025年 240卷

作者： Zhou, Shuangxi Yang, Dan Zhang, Ziyu Zhang, Jinwen Qu, Fulin Punetha, Piyush Li, Wengui Li, Ning Guangzhou Maritime Univ Sch Civil & Engn Management Guangzhou 510725 Peoples R China East China Jiaotong Univ Sch Civil Engn & Architecture Nanchang 330013 Peoples R China Univ New South Wales Ctr Infrastruct Engn & Safety Sch Civil & Environm Engn Sydney NSW 2052 Australia Univ Technol Sydney Sch Civil & Environm Engn Ultimo NSW 2007 Australia Univ Manchester Dept Solids & Struct Manchester M13 9PL England

To enhance the safety and comfort of vehicle travel, detecting pavement cracks is a critical task in road management. This article introduces an advanced single-stage target detection method utilizing the yolov5s algorithm to enhance real-time performance and accuracy. Initially, Squeeze-and-Excitation Networks are integrated into the model to facilitate self-learning for improved crack characterization. Subsequently, anchors computed through the K-means clustering algorithm are closely aligned with the fracture dataset, achieving an adaptation rate of 99.9 % and enhancing the recall rate of the model. Furthermore, the inclusion of the SimSPPF module from yolov6 diminishes memory usage and expedites detection speed. By replacing the original nearest up- sampling method with transposed convolution, optimization of up-sampling for crack datasets is achieved. Performance assessments reveal that the refined yolov5s algorithm attains an F1 score of 91 %, a mean Average Precision (mAP) of 93.6 %, and a 1.54 % increase in frames per second (fps) for pavement crack detection. This enhancement in detection technology signifies a substantial advancement in the maintenance and longevity of road infrastructure.

关键词： Road maintenance Crack detection yolo algorithm AI integration Advanced image processing

来源：评论

学校读者我要写书评

暂无评论

Detection of abandoned objects based on yolov9 and background differencing

引用

SIGNAL IMAGE AND VIDEO PROCESSING 2025年第1期19卷 1-8页

作者： Song, Huajun Wang, Jinbo Zhang, Yunze China Univ Petr East China Coll Oceanog & Space Informat Changjiang West Rd Qingdao 266580 Shandong Peoples R China

Abandoned objects on highways can cause serious traffic accidents. Abandoned objects detection algorithms are of great significance in reducing traffic accidents and improving the efficiency of highway management. Existing traditional image processing detection algorithms are difficult to detect small targets in the complex and changing highway environment. And the current deep learning methods are limited by the number of samples and the influence of vehicle occlusion resulting in lower detection accuracy. Aiming at the above problems, a abandoned object detection algorithm for highway is designed by combining the background modeling algorithm MOG2 and the yolov9 detector. Utilizing the difference of the two algorithms on the recognition mechanism of abandoned objects, the results of the yolov9 detector are used as a guide to make a comprehensive judgment on the low confidence targets combined with the iterative difference results of MOG2. The actual test results on highways show that the abandoned objects detection algorithm based on yolov9 and MOG2 has an improved precision and recall rate for abandoned objects detection compared to a single detection algorithm, with an average precision of 86.7%\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\%$$\end{document} and a recall rate of 89.6%\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\%$$\end{document}. The algorithm effectively complements the existing abandoned objects detection algorithms and has engineering practical value.

关键词： Abandoned object detection Computer vision yolo algorithm Gaussian mixture model

来源：评论

学校读者我要写书评

暂无评论

Enhancing Bolt Object Detection via AIGC-Driven Data Augmentation for Automated Construction Inspection

引用

BUILDINGS 2025年第5期15卷 819-819页

作者： Wu, Jie Han, Beilin Zhang, Yihang Huang, Chuyue Qiu, Shengqiang Feng, Wang Liu, Zhiwei Zou, Chao Wuhan Polytech Univ Sch Civil Engn & Architecture Wuhan 430023 Peoples R China Guangdong Prov Key Lab Intelligent Disaster Preven Dongguan 523808 Peoples R China Wuhan Univ Technol Sch Transportat & Logist Engn Wuhan 432063 Peoples R China Cent & Southern China Municipal Engn Design & Res Wuhan 430010 Peoples R China Guangdong Univ Technol Sch Civil & Transportat Engn Guangzhou 510006 Peoples R China

In the engineering domain, the detection of damage in high-strength bolts is critical for ensuring the safe and reliable operation of equipment. Traditional manual inspection methods are not only inefficient but also susceptible to human error. This paper proposes an automated bolt damage identification method leveraging AIGC (Artificial Intelligence Generated Content) technology and object detection algorithms. Specifically, we introduce the application of AIGC in image generation, focusing on the Stable Diffusion model. Given that the quality of bolt images generated directly by the Stable Diffusion model is suboptimal, we employ the LoRA fine-tuning technique to enhance the model, thereby generating a high-quality dataset of bolt images. This dataset is then used to train the yolo (You Only Look Once) object detection algorithm, demonstrating significant improvements in both accuracy and recall for bolt damage recognition. Experimental results show that the LoRA fine-tuned Stable Diffusion model significantly enhances the performance of the yolo algorithm, providing an efficient and accurate solution for automated bolt damage detection. Future work will concentrate on further optimizing the model to improve its robustness and real-time performance, thereby better meeting the demands of practical industrial applications.

关键词： high-performance bolts damage detection automated detection AIGC technology yolo algorithm

来源：评论

学校读者我要写书评

暂无评论

Effects of Nocturnal Grazing Behavior and Stock Density on Cattle Dung in Sloped Rotational Grazing

引用

RANGELAND ECOLOGY & MANAGEMENT 2025年 99卷 18-25页

作者： Nakashima, Naohisa Kawamura, Kensuke Hirata, Masahiro Muneoka, Toshimi Kimura, Masato Shinkai, Moeka Obihiro Univ Agr & Vet Med Dept Agroenvironm Sci Obihiro Hokkaido 0808555 Japan Obihiro Univ Agr & Vet Med Dept Human Sci Obihiro Hokkaido 0808555 Japan AK Konzern Ltd Sapporo Hokkaido 0460202 Japan

In this study, we investigated the effects of nocturnal variation in grazing behavior and stock density on the distribution of cattle dung in a sloped rotational grazing system in Japan. The specific objectives were to determine the influence of watering sites on dung distribution at night and the relationship between stock density and dung distribution. The study was conducted on a 2.56-ha steeply sloped pasture that was divided into four sub-paddocks for rotational grazing. Drone imagery and the yolo deep learning algorithm were used to automate the mapping of more than 400 cattle dung pats per grazing session, with a mean average precision of 0.75. The yolo algorithm demonstrated high accuracy in identifying dung that was up to 4 days old. Our analyses revealed that the distance to a watering site influenced fecal distribution during the day, but had a weaker and insignificant effect at night. With low-density grazing, dung pats showed a clumped distribution across the entire pasture, whereas at higher stock density the distribution approached randomness. The formation of four dung hotspots was observed in the pasture, although these spots were not consistently used throughout subsequent grazing periods. The cattle tended to migrate from more elevated foraging areas to more level grazing areas for rest, resulting in a loose connection between foraging and resting sites, and this behavior may have a bearing on the lack of successive utilization of these fecal hotspots. Subsequent research must examine whether these patterns persist across seasonal variations and different grazing lands. Despite this need for validation, our findings can provide insights for guiding sustainable nutrient redistribution and grassland management on sloped terrain through the management of stock density and grazing schedules. (c) 2024 The Society for Range Management. Published by Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar

关键词： Automated identification Cattle dung Pasturing Sloped pasture Stock density yolo algorithm

来源：评论

学校读者我要写书评

暂无评论

Analysis, Design and Implementation of a Ripe Mango Detection Program in Burkina Faso 15th

Analysis, Design and Implementation of a Ripe Mango Detectio...

引用

15th International Conference on e Infrastructure and e Services for Developing Countries

作者： Bikienga, Moustapha Tougma, Roland Manegaouinde Ouedraogo, Soumaila Norbert ZONGO Univ Avce Maurice Yameogo BP376 Koudougou Burkina Faso Joseph KI ZERBO Univ Ouagadougou CFX2 7R6 Burkina Faso

ISBN: (纸本)9783031815720;9783031815737

The analysis, concept, and implementation of a computer program capable of detecting ripe mangoes are at the core of this study. Traditional methods are often faced with calibration errors. Recently, deep learning has shown promising performance in visually guided agricultural applications. Faced with these constraints, it is necessary to establish an automatic system for robust and efficient detection of mangoes in orchards. In this study, a fast implementation system of a mango detector, distinguishing between ripe and unripe mangoes based on deep learning using the yolov5 algorithm, was developed. From a simple photo, the algorithm detects and counts the number of mangoes on a tree. This artificial intelligence system (deep neural network) was trained on a dataset of over 500 annotated mango images. Experimental results show that the algorithm achieves 98% precision, 98% recall, and an F1-score of 98%. This satisfactory precision in mango detection offers significant advantages in terms of efficiency and accuracy compared to traditional methods. However, it should be noted that our system has certain limitations. Nevertheless, our study demonstrates promising results in the field of ripe mango detection.

关键词： Deep learning yolo algorithm Mangoes Performance metrics Computer vision

来源：评论

学校读者我要写书评

暂无评论

AI-Assisted Web Application for Leukocyte Abnormality Counting with yolov11 and Smartphone Microscopy

引用

IEEE Access 2025年 13卷 89079-89107页

作者： Kasamsumran, Natthakorn Ittichaiwong, Piyalitt Chinudomporn, Chutinart Veerakanjana, Kanyakorn Karoopongse, Ekapun Pora, Wanchalerm Chulalongkorn University Faculty of Engineering Department of Electrical Engineering Bangkok Thailand Mahidol University Siriraj Informatics and Data Innovation Center Faculty of Medicine Siriraj Hospital Bangkok Thailand King's College London Faculty of Life Sciences & Medicine Department of Cancer Imaging School of Biomedical Engineering & Imaging Sciences London United Kingdom Mahidol University Faculty of Medicine Ramathibodi Hospital Bangkok Thailand King's College London Social Genetic & Developmental Psychiatry Centre Institute of Psychiatry Psychology & Neuroscience London United Kingdom Mahidol University Division of Hematology Department of Internal Medicine Faculty of Medicine Siriraj Hospital Bangkok Thailand

Accurate and timely white blood cell (WBC) analysis is crucial for diagnosing hematological disorders, often requiring microscopic examination of peripheral blood smears (PBS). While manual counting by trained specialists is considered the gold standard, it is time-consuming and impractical in resource-limited settings. Automated cell counters can misclassify similitude or immature cells, hindering accurate diagnosis. To address these limitations, we propose an AI-powered web application that utilizes yolov11 with enhanced small object detection capabilities, enabled by integrating our C3k2-Conv blocks, an architecture inspired by C3k2. Our model, trained on eleven WBC classes and nucleated red blood cells (NRBCs), achieves an impressive mean average precision (mAP@0.5) of approximately 0.9000 on validation and unseen test sets, demonstrating a performance comparable to human specialists in identifying and quantifying WBCs. Furthermore, our research demonstrates that providing general practitioners and medical students with PBS images annotated by our AI model significantly improves their counting accuracy and reduces the time spent on manual counting. Our web application, Myelosoft, allows clinicians to upload smartphone-captured PBS images for rapid and automated analysis. The system provides comprehensive differential counts for 11 WBC classes, including atypical lymphocytes, band neutrophils, basophils, blasts, eosinophils, lymphocytes, metamyelocytes, monocytes, myelocytes, promyelocytes, and segmented neutrophils, as well as NRBCs. This real-time analysis facilitates timely diagnosis and treatment, potentially reducing risks associated with delayed interventions. Our approach offers a robust and accessible solution for improving hematologic treatment, especially in resource-constrained environments. © 2013 IEEE.

关键词： Human vs. Artificial Intelligence Leukocyte Detection Microscopy Imaging Object Detection Peripheral Blood Smear Web Application White Blood Cell yolo algorithm

来源：评论

学校读者我要写书评

暂无评论

The improvement in obstacle detection in autonomous vehicles using yolo non-maximum suppression fuzzy algorithm

引用

JOURNAL OF SUPERCOMPUTING 2021年第11期77卷 13421-13446页

作者： Zaghari, Nayereh Fathy, Mahmood Jameii, Seyed Mahdi Shahverdy, Mohammad Islamic Azad Univ Dept Comp Engn Shahr E Qods Branch Tehran Iran Inst Res Fundamental Sci Sch Comp Sci Tehran Iran IRAN Univ Sci & Technol IUST Dept Comp Engn Tehran Iran

Numerous changes in algorithms have been observed by object detection to enhance both speed and accuracy. In this research, we present a method to improve the behavioral clone of self-driving cars. Thus, we first create a collection of videos and information required for safe driving on different routes and conditions. The detection of obstacles is done with the proposed algorithm called "yolo non-maximum suppression fuzzy algorithm, which performs the driver reaction to obstacles with greater accuracy and more speed than the obstacles detection algorithms using the designed framework. The network is trained by the driver's performance, and hence, the output used to control the vehicle is obtained. The non-maximum suppression algorithm plays an essential role in object detection and tracking. An effective hybrid method of fuzzy and NMS algorithms is provided in this paper to improve the problem mentioned. The proposed method improves the average accuracy of the detection network. The performance of the designed algorithm was examined using two different types of KITTI data and the data collected using the personal vehicle and the data we gathered. The proposed algorithm was assessed with evaluation accuracy criteria, which revealed that the method has a higher speed (above 64.41%), a lower FPR (below 6.89%), and a lower FNR (below 3.95%) compared with the baseline yolov3 model. According to the loss function, the accuracy rate of the network performance is 95%, implying that we have achieved good results.

关键词： Self-driving Object detection yolo algorithm Non-maximum suppression algorithm Fuzzy Convolutional neural network (CNN)

来源：评论

学校读者我要写书评

暂无评论

Enhanced yolov8 Integrated Brain-Inspired Attention Mechanisms for Weed Detection 17th

Enhanced YOLOv8 Integrated Brain-Inspired Attention Mechanis...

引用

17th International Conference on Intelligent Robotics and Applications

作者： Wang, Maosheng Fang, Yuan Univ Bristol Bristol BS8 1QU Avon England Univ Chinese Acad Sci Beijing 100190 Peoples R China

ISBN: (纸本)9789819607884;9789819607891

This study explores the application of yolov8 in agricultural weed detection and compares the performance of other versions of yolo, as well as the yolov8 model incorporating brain-inspired attention mechanisms, in terms of weed detection. The comparison shows that the yolov8 model enhances detection accuracy and speed through its advanced network architecture and image processing techniques. The incorporation of attention mechanisms further improves its detection capabilities by focusing on key features within images, optimizing the model's ability to discern subtle differences between weeds and crops. This improvement further increases its efficiency in complex agricultural environments. Experimental validation demonstrates that yolov8 significantly surpasses previous models in terms of precision and reliability, supporting its integration into precision agriculture. This research emphasizes the importance of adopting improved intelligent technologies like yolov8 to enhance the efficiency and sustainability of agriculture.

关键词： Weed detection yolo algorithm Brain-inspired model Attention mechanism Smart agriculture

来源：评论

学校读者我要写书评

暂无评论

Recognition of big mammal species in airborne thermal imaging based on yolo V5 algorithm

引用

Integrative Zoology 2023年第2期18卷 333-352页

作者： Yonghua XIE Jueze JIANG Heng BAO Penghui ZHAI Yue ZHAO Xiaoyu ZHOU Guangshun JIANG College of Mechanical and Electrical Engineering Northeast Forestry UniversityHarbinChina Feline Research Center of National Forestry and Grassland Administration College of Wildlife and Protected AreaNortheast Forestry UniversityHarbinChina Harbin Siberian Tiger Pk HarbinChina Research Center for Northeast Asia Biodiversity Conservation HarbinChina

Unmanned aerial vehicle(UAV)technology,artiﬁcial intelligence,and the relevant hardware can be used for monitoring wild ***,existing methods have several ***,this study explored the monitoring and protection of Amur tigers and their main prey species using images from UAVs by optimizing the algorithm models with respect to accuracy,model size,recognition speed,and elimination of environmental *** imaging data were collected from 2000 pictures with a thermal imaging lens on a DJI M300RTK UAV at the Hanma National Nature Reserve in the Greater Khingan Mountains in Inner Mongolia,Wangqing National Nature Reserve in Jilin Province,and Siberian Tiger Park in Heilongjiang *** yolo V5s al-gorithm was applied to recognize the animals in the *** accuracy rate was 94.1%,and the size of the model weight(total weight of each model layer trained with the training set)was 14.8 *** authors improved the structures and parameters of the yolo V5s *** a result,the recognition accuracy rate became 96%,and the model weight was 9.3 *** accuracy rate increased by 1.9%,the model weight decreased by 37.2%from 14.8 MB to 9.3 MB,and the recognition time of a single picture was shortened by 34.4%from 0.032 to 0.021 *** not only increases the recognition accuracy but also effectively lowers the hardware requirements that the algorithm relies on,which provides a lightweight fast recognition method for UAV-based edge computing and online investigation of wild animals.

关键词： online investigation of wild animals recognition research UAV-based thermal imaging yolo algorithm

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：