环形编码标志点鲁棒识别研究  

Robust Recognition of Circular Coded Markers

作　　者：王洪伟 亢太波 邓海涛 陈庭 WANG Hongwei;KANG Taibo;DENG Haitao;CHEN Ting(Wuhan Municipal Road and Bridge Co.Ltd,Wuhan 430015,China;School of Geodesy and Geomatics,Wuhan University,Wuhan 430079,China;Key Laboratory of Geospace Environment and Geodesy,Ministry of Education,Wuhan University,Wuhan 430079,China)

机构地区：[1]武汉市政路桥有限公司,湖北武汉430015 [2]武汉大学测绘学院,湖北武汉430079 [3]武汉大学地球空间环境与大地测量教育部重点实验室,湖北武汉430079

出　　处：《武汉大学学报（信息科学版）》2024年第9期1734-1740,共7页Geomatics and Information Science of Wuhan University

基　　金：国家重点研发计划(2018YFC1503605)。

摘　　要：环形编码标志点在近景摄影测量中应用广泛,具有平移、旋转、缩放无关的优良特性。环形编码标志点的鲁棒识别与解码对后续三维重建工作有着重要意义。针对杂乱背景、不均匀光照等复杂环境下的编码标志点识别,提出了一种两步法的轮廓筛选策略,首先基于轮廓长度和圆度进行粗筛,然后基于灰度分布统计分析进行精细筛选。实验结果表明,当拍摄角度小于75°且编码点成像大于30像素时,能够保证较高的识别准确率与解码准确率;在复杂环境下,识别正确率和解码正确率分别为99.53%和100%,误识别率为0%。实验证明,所提方法有效降低了误识别率,在复杂环境下具有良好的识别效果。Objectives:The circular coded marker is widely used in close-range photogrammetry,and has excellent characteristics that are independent of translation,rotation,and scaling.The robust recognition and decoding of the circular coded markers are of great significance to the subsequent 3D reconstruction.Methods:Aiming at the identification of coded marks in complex environments such as cluttered backgrounds and uneven lighting,a two-step contour screening strategy is proposed.First,coarse screening is performed based on the contour length and roundness,and then fine screening is performed based on statistical analysis on gray-scale distribution.Results:The experimental results show that when the shooting angle is less than 75°and the size of the code point is greater than 30 pixels,the proposed method can guarantee high accuracy in recognition and decoding.In a complex environment,the recognition accuracy and decoding accuracy are 99.53%and 100%,respectively,and the false recognition rate is 0%.Conclusions:The proposed method can effectively reduce the false recognition rate and have a good recognition effect in a complex environment.

关 键 词：近景摄影测量 编码标志点 目标识别 图像处理 轮廓筛选 

分 类 号：P228[天文地球—大地测量学与测量工程]