现代电子技术2024，Vol.47Issue(20)：1-7,7.DOI:10.16652/j.issn.1004-373x.2024.20.001
结合YOLOv8与双目测距算法的水面漂浮垃圾检测定位系统设计
Design of water surface floating garbage detection and positioning system combining YOLOv8 and binocular ranging algorithm
摘要
Abstract
In order to solve the problem of poor location and recognition of garbage target of automatic garbage recovery ship in outdoor waters,a garbage identification and location system of floating garbage recovery ship is proposed,which combines YOLOv8 and binocular ranging algorithm.The system is mainly consisted of three parts:camera,upper computer visual processing unit,and lower computer control unit.After visual identification and classification of garbage in the water environment,location and ranging are carried out to realize the location and identification of garbage in the water area.After positioning and recognition,the machine is controlled to collect garbage.Binocular camera is used to obtain images,Jetson nano embedded chip is used as the host computer main control chip,the latest deep learning model YOLOv8 is used to extract and identify surface garbage,and binocular distance is obtained by means of SGBM algorithm.The distance and angle information measured by the upper computer are sent to the lower computer Arduino control board by the serial communication to control the ship's turning and navigation.The testing results show that the recognition result of the collection device is stable,the accuracy rate can reach 90.5%,the ranging result is accurate,the accuracy can reach centimeter level,and the automatic collection effect of the control device can be achieved.关键词
水面漂浮垃圾/目标定位/垃圾识别/YOLOv8/双目测距算法/视觉检测/自动收集Key words
water surface floating garbage/target positioning/garbage identification/YOLOv8/binocular ranging algorithm/visual detection/automatic collection分类
信息技术与安全科学引用本文复制引用
何君尧,王文胜,韩宜航..结合YOLOv8与双目测距算法的水面漂浮垃圾检测定位系统设计[J].现代电子技术,2024,47(20):1-7,7.基金项目
国家重点研发计划课题(2020YFB1713205) （2020YFB1713205）
北京市教育委员会科研计划项目资助(KM202411232023) （KM202411232023）
2024年北京信息科技大学"青年骨干教师"支持计划(YBT202403) （YBT202403）
