| 程 亮,吴兴辉,江云华,苏 雄,吴佳晓,周 辉,丁美有,何赟泽.基于无人船视觉的水域人员类别识别算法[J].电子测量与仪器学报,2022,36(8):43-51 |
| 基于无人船视觉的水域人员类别识别算法 |
| Person category identification algorithm in water environmentbased on unmanned ship vision |
| |
| DOI： |
| 中文关键词: 水面无人船 水域人员识别 YOLO v5 网络部署 |
| 英文关键词:USV person identification in water environment YOLO v5 network deployment |
| 基金项目:湖南省自然科学基金重大项目、珠海云洲智能科技有限公司委托课题项目资助 |
|
|
| Author | Institution |
| Cheng Liang | 1. School of Ocean Engineering, Jiangsu Ocean University, 2. Zhuhai Yunzhou Intelligent Technology Co. , Ltd. |
| Wu Xinghui | 3. College of Electrical and Information Engineering, Hunan University |
| Jiang Yunhua | 2. Zhuhai Yunzhou Intelligent Technology Co. , Ltd. |
| Su Xiong | 2. Zhuhai Yunzhou Intelligent Technology Co. , Ltd. |
| Wu Jiaxiao | 2. Zhuhai Yunzhou Intelligent Technology Co. , Ltd. |
| Zhou Hui | 3. College of Electrical and Information Engineering, Hunan University |
| Ding Meiyou | 3. College of Electrical and Information Engineering, Hunan University |
| He Yunze | 3. College of Electrical and Information Engineering, Hunan University |
|
| 摘要点击次数: 1570 |
| 全文下载次数: 1098 |
| 中文摘要: |
| 针对水域环境下人员识别,提出了一种基于水面无人船(unmanned surface ship,USV)视觉传感器的水域人员类别识别
算法。 依照数据采集与模型更新流程,将采集到的视频数据进行数据清洗与标记后,创建人员类别数据集 39 959 张图片,7 个
类别;实践了基于深度学习方法下主流目标检测网络 YOLO v5,并针对水域环境场景特点,提出基于 YOLO v5 的人员类别识别
算法;将人员类别识别算法部署到边缘计算平台,实现算法在无人船上的实时应用。 算法在人员类别识别数据集上达到了平均
精度 86%,在无人船实测中实现了每秒处理 38 帧的人员类别识别实时性表现。 |
| 英文摘要: |
| To achieve person recognition in water environment, a person category identification algorithm based on vision sensors on
unmanned surface ship (USV) is proposed. Firstly, base on the data acquisition and model update workflow, a person category dataset
of 39 959 pictures and 7 categories is created after data cleaning and labeling on original videos. Secondly, YOLO v5, the mainstream
object detection network in the field of deep learning method, is practiced, and an improved person category identification algorithm
based on YOLO v5 is proposed according to the characteristics of water environment scenes. Thirdly, the algorithm is deployed to the
edge computing platform to realize the real-time use of the algorithm on the unmanned ship. The algorithm achieves an average accuracy
of 86% on our dataset and achieves real-time performance of processing 38 frames per second with accurate person recognition in the
unmanned ship test. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
