AUV自主巡航与色块识别的典型方法与应用前景  

Typical approaches and application prospects for autonomous robot navigation and color-block identification of AUV

作　　者：王慧[1] 李家锐 黄栋 黄平洋 许时镇 孔祥峰 李慎德[1] 陈春雷[1] Wang Hui;Li Jiarui;Huang Dong;Huang Pingyang;Xu Shizhen;Kong Xiangfeng;Li Shende;Chen Chunlei(Guangdong Ocean University,Zhanjiang 524088,China)

机构地区：[1]广东海洋大学,广东湛江524088

出　　处：《无线互联科技》2020年第15期98-102,110,共6页Wireless Internet Technology

基　　金：广东海洋大学创新创业训练计划项目,项目编号:CXXL2019262;广东省教育厅校企合作协同育人“粤嵌科技校企合作——创新创业教育改革”项目,项目编号:PROJ1112628756155076608。

摘　　要：文章讨论了无缆水下机器人(AUV)在复杂的水环境中用同步自主定位与地图构建和路径规划如何实现自主巡航的问题,分析与比较了基于树莓派的OpenCV、神经网络的Tensorflow以及STM32的OpenMV这3种典型的色块识别方法的优劣。This article has a discussion on the issue how to accomplish the aim of autonomous robot navigation with simultaneous localization and map construction and path planning in the complex water environment by autonomous underwater vehicle,at the same time,the pros and cons of three typical approaches to color-block identification are analyzed and compared,including OpenCV based on raspberry-pi,Tensorflow provided with neural network and OpenMV building on STM32.

关 键 词：自主巡航 色块识别 无缆水下机器人 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置] U674.941[自动化与计算机技术—控制科学与工程]