电子测量与仪器学报

吴兴辉,何赟泽,周辉,程亮,丁美有.改进 YOLO v7 算法下的监控水域环境人员识别研究[J].电子测量与仪器学报,2023,37(5):20-27

改进 YOLO v7 算法下的监控水域环境人员识别研究

Research on the personnel recognition in monitored water areabased on improved YOLO v7 algorithm

DOI：

英文关键词:personnel recognition YOLO-WA attention mechanism

基金项目:湖南省自然科学基金杰出青年基金项目(2022JJ10017)、珠海云洲智能科技有限公司委托课题(H202191400377)项目资助

作者	单位
吴兴辉	1. 湖南大学电气与信息工程学院
何赟泽	1. 湖南大学电气与信息工程学院
周辉	1. 湖南大学电气与信息工程学院
程亮	2. 江苏海洋大学海洋工程学院连云港 222005; 3. 珠海云洲智能科技股份有限公司
丁美有	1. 湖南大学电气与信息工程学院

Author	Institution
Wu Xinghui	1. College of Electrical and Information Engineering, Hunan University
He Yunze	1. College of Electrical and Information Engineering, Hunan University
Zhou Hui	1. College of Electrical and Information Engineering, Hunan University
Cheng Liang	2. School of Ocean Engineering, Jiangsu Ocean University, 3. Zhuhai Yunzhou Intelligent Technology Co. , Ltd.
Ding Meiyou	1. College of Electrical and Information Engineering, Hunan University

摘要点击次数: 465

全文下载次数: 664

中文摘要:

基于水域监控系统智能化的发展需求,提出了一种监控水域环境下人员识别算法。在水域场景数据采集、数据清洗与标记后,自主构建了一套监控水域场景下的人员类别数据集 YZ-Water4,共 8 092 张图片和 24 011 个标签。基于目标检测算法 YOLO v7 的性能基础,针对水域场景特点,提出了适用于水域环境的目标检测算法 YOLO-WA( you only look once-water area)。首先,使用更适合视觉任务的 FReLU 激活函数取代 YOLO v7 算法中激活函数;其次将注意力机制融合到算法网络骨架中,提升算法的特征提取能力;最后,选择 SIOU 损失函数替换 YOLO v7 算法中的 CIOU 损失函数以优化算法训练过程。实验结果表明, YOLO-WA 与原算法相比,在水域人员类别数据集上识别精确率由 82. 3%提升到 86. 9%,召回率由 92. 0%提升到 92. 8%,平均精度从 88. 4%提高到 90. 6%,检测速度达到了 85 fps,满足实时运行的精度与速度要求。

英文摘要:

Based on the development demand of intelligent water area monitoring system, a personnel recognition algorithm for monitored water area is proposed. After data collection of the water area scene, data cleaning and labeling, a personnel category dataset YZ-Water4 under the monitored water area scene was independently constructed, with a total of 8 092 images and 24 011 tags. Based on the performance of the object detection algorithm YOLO v7 and the characteristics of the water area scene, object detection algorithm YOLOWA (you only look once-water area) for water environment is proposed. First, the FReLU activation function which is proposed for visual tasks is used to replace the activation function in YOLO v7 algorithm. Secondly, the attention mechanism is integrated into algorithm to improve the feature extraction ability of the algorithm. Finally, SIOU loss function is chosen to replace CIOU loss function in YOLO v7 algorithm to optimize the training process. The experimental results show that compared with the original algorithm, YOLO-WA has increased the precision rate from 82. 3% to 86. 9%, recall rate from 92. 0% to 92. 8%, mean average precision from 88. 4% to 90. 6%, and the processing speed is 85 frame per second, meeting the accuracy and speed requirements of real-time run.

查看全文查看/发表评论下载PDF阅读器