{"title":"Research on Improved Algorithm for Small Object Detection in Intelligent Surveillance Video based on YOLOv7","authors":"Zhiwei Wang, Min Wang","doi":"10.54097/ehvf7754","DOIUrl":null,"url":null,"abstract":"In order to address the issue of small objects being difficult to detect effectively in intelligent surveillance videos, this study proposes an improved scheme for the YOLOv7-tiny algorithm. This scheme integrates the Convolutional Block Attention Module (CBAM) into YOLOv7-tiny, effectively enhancing the model's feature extraction and small object detection capabilities in complex backgrounds, thereby improving the overall detection precision. Experimental evaluations indicate that the improved algorithm shows enhanced performance in specific small object detection tasks, achieving an accuracy of 85.6%, a recall rate of 85.2%, and a mean average precision (mAP) of 90.2%. These results demonstrate the effectiveness and practical value of the improved scheme in enhancing the performance of YOLOv7-tiny in small object detection tasks.","PeriodicalId":504530,"journal":{"name":"Frontiers in Computing and Intelligent Systems","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Frontiers in Computing and Intelligent Systems","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.54097/ehvf7754","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
In order to address the issue of small objects being difficult to detect effectively in intelligent surveillance videos, this study proposes an improved scheme for the YOLOv7-tiny algorithm. This scheme integrates the Convolutional Block Attention Module (CBAM) into YOLOv7-tiny, effectively enhancing the model's feature extraction and small object detection capabilities in complex backgrounds, thereby improving the overall detection precision. Experimental evaluations indicate that the improved algorithm shows enhanced performance in specific small object detection tasks, achieving an accuracy of 85.6%, a recall rate of 85.2%, and a mean average precision (mAP) of 90.2%. These results demonstrate the effectiveness and practical value of the improved scheme in enhancing the performance of YOLOv7-tiny in small object detection tasks.