摘要
针对现有算法在解决无人机(UAV)协同跟踪过程中的实时视觉跟踪性能不足等问题,利用单目视觉技术在自主检测和跟踪运算处理强度,避障路径决策等提出了一种基于机器视觉自主检测和协同跟踪算法。该算法利用模板匹配(TM)和形态滤波(MF)混合算法,确保环境照明条件和背景变化的性能稳定下,通过限制覆盖的搜索区域来减少计算量,提高计算效率并优化检测策略,减少错误报警和遗漏检测。通过特定无人机仿真试验数据分析,验证检测和跟踪算法的可行性,表明无人机在特定环境下的跟踪检测性能得到显著提高。
In order to solve the problems of insufficient real-time visual tracking performance of existing algorithms in the process of unmanned aerial vehicle(UAV)cooperative tracking,this paper uses monocular vision technology in autonomous detection and tracking calculation processing strength,obstacle avoidance path decision,etc.A machine vision detection and collaborative tracking algorithm is presented.The algorithm uses a hybrid algorithm of template matching(TM)and morphological filtering(MF)to ensure stable performance under environmental lighting conditions and background changes,reducing the amount of calculation by limiting the search area covered,improving calculation efficiency and optimizing detection strategies to reduce errors Alarm and omission detection.Through the analysis of specific UAV simulation test data to verify the feasibility of the detection and tracking algorithm,it shows that the UAV’ss tracking and detection performance in a specific environment has been significantly improved.
作者
陆渊章
戴红霞
胡莹
王志亮
LU Yuanzhang;DAI Hongxia;HU Ying;WANG Zhiliang(School of Electronic Information Engineering,Jiangsu Vocational College of Information Technology,Wuxi Jiangsu 214153,China;School of Information Science and Engineering,Southeast University,Nanjing Jiangsu 212003,China;Key Laboratory of ASIC Design of Jiangsu Province,Nantong University,Nantong Jiangsu 226007,China)
出处
《电子器件》
CAS
北大核心
2020年第5期1096-1099,共4页
Chinese Journal of Electron Devices
基金
国家自然科学基金项目(61801435,61801170)
江苏高校自然科学基金研究面上项目(16KJB510008)
江苏省高水平骨干专业建设项目(苏教高[2017]17号)
江苏省高等职业教育产教融合集成平台项目(苏教职函[2019]26号)
江苏省专用集成电路设计重点实验室开放课题基金项目(20201012)。
关键词
无人机
视觉跟踪
形态滤波
视线估计
UAV(unmanned aerial vehicle)
visual tracking
morphological filtering
LOS(light of sight)estimation