YOLO: A Competitive Analysis of Modern Object Detection Algorithms for Road Defects Detection Using Drone Images

IF 1.2 Q3 MULTIDISCIPLINARY SCIENCES

Baghdad Science Journal Pub Date : 2023-11-19 DOI:10.21123/bsj.2023.9027

Amit Hasan Sadhin, Siti Zaiton Mohd Hashim, Hussein Samma, Nurulaqilla Khamis

{"title":"YOLO: A Competitive Analysis of Modern Object Detection Algorithms for Road Defects Detection Using Drone Images","authors":"Amit Hasan Sadhin, Siti Zaiton Mohd Hashim, Hussein Samma, Nurulaqilla Khamis","doi":"10.21123/bsj.2023.9027","DOIUrl":null,"url":null,"abstract":"Efficient identification of road defects is a critical concern for road safety and infrastructure upkeep. This research employs drone-captured imagery and advanced object detection algorithms to expedite defect recognition, with a specific focus on determining the optimal algorithm for prompt and precise detection. The importance of timely road defect detection, crucial for mitigating potential hazards, remains central. A comprehensive comparative analysis of contemporary object detection algorithms, encompassing YOLOv5s, YOLOv5m, YOLOv5l, YOLOv5x, and YOLOv7. The results of this study highlight YOLOv7 as the most efficient, with a notable mAP of 68.3%, closely followed by YOLOv5l (66.8%), YOLOv5m (66.3%), YOLOv5x (66%), and YOLOv5s (63%). The integration of drone-derived imagery, capturing distinct gradients, significantly enhances defect detection accuracy. Beyond road safety, this study offers valuable insights to computer vision and machine learning practitioners. By bridging technological innovation with practical implementation, it holds potential to advance road safety and transportation infrastructure quality and the use of revolutionary drone technology.","PeriodicalId":8687,"journal":{"name":"Baghdad Science Journal","volume":"132 1","pages":""},"PeriodicalIF":1.2000,"publicationDate":"2023-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Baghdad Science Journal","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.21123/bsj.2023.9027","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"MULTIDISCIPLINARY SCIENCES","Score":null,"Total":0}

引用次数: 0

Abstract

Efficient identification of road defects is a critical concern for road safety and infrastructure upkeep. This research employs drone-captured imagery and advanced object detection algorithms to expedite defect recognition, with a specific focus on determining the optimal algorithm for prompt and precise detection. The importance of timely road defect detection, crucial for mitigating potential hazards, remains central. A comprehensive comparative analysis of contemporary object detection algorithms, encompassing YOLOv5s, YOLOv5m, YOLOv5l, YOLOv5x, and YOLOv7. The results of this study highlight YOLOv7 as the most efficient, with a notable mAP of 68.3%, closely followed by YOLOv5l (66.8%), YOLOv5m (66.3%), YOLOv5x (66%), and YOLOv5s (63%). The integration of drone-derived imagery, capturing distinct gradients, significantly enhances defect detection accuracy. Beyond road safety, this study offers valuable insights to computer vision and machine learning practitioners. By bridging technological innovation with practical implementation, it holds potential to advance road safety and transportation infrastructure quality and the use of revolutionary drone technology.

查看原文本刊更多论文

YOLO：利用无人机图像检测道路缺陷的现代物体检测算法竞争分析

有效识别道路缺陷是道路安全和基础设施维护的关键问题。本研究利用无人机捕捉的图像和先进的物体检测算法来加快缺陷识别，重点是确定最佳算法，以实现及时、精确的检测。及时发现道路缺陷对减少潜在危险至关重要，这一点仍然是核心问题。对当代物体检测算法进行了全面的比较分析，包括 YOLOv5s、YOLOv5m、YOLOv5l、YOLOv5x 和 YOLOv7。研究结果表明，YOLOv7 是最高效的算法，mAP 值高达 68.3%，紧随其后的是 YOLOv5l（66.8%）、YOLOv5m（66.3%）、YOLOv5x（66%）和 YOLOv5s（63%）。整合无人机获取的图像，捕捉明显的梯度，大大提高了缺陷检测的准确性。除道路安全外，本研究还为计算机视觉和机器学习从业人员提供了宝贵的见解。通过将技术创新与实际应用相结合，该研究有望促进道路安全、交通基础设施质量以及革命性无人机技术的使用。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Baghdad Science Journal MULTIDISCIPLINARY SCIENCES-

CiteScore

2.00

自引率

50.00%

发文量

102

审稿时长

24 weeks

期刊介绍： The journal publishes academic and applied papers dealing with recent topics and scientific concepts. Papers considered for publication in biology, chemistry, computer sciences, physics, and mathematics. Accepted papers will be freely downloaded by professors, researchers, instructors, students, and interested workers. ( Open Access) Published Papers are registered and indexed in the universal libraries.