质心
亚像素渲染
灰度
像素
人工智能
光学
计算机视觉
摄影测量学
计算机科学
均方误差
电荷耦合器件
物理
数学
统计
作者
Kaifeng Ma,Guiping Huang,Junzhen Meng
出处
期刊:Applied Optics
[Optica Publishing Group]
日期:2024-04-16
卷期号:63 (14): 3965-3965
被引量:2
摘要
Depending on the size, shape, and gray intensity distribution of charge-coupled device (CCD) imaging retro-reflective targets (RRTs) in close-range photogrammetry, and based on conventional grayscale centroiding, this paper proposes grayscale threshold variable-index weighted centroiding (GTVIWC). The centroid location accuracy of CCD imaging RRTs was analyzed and compared using simulated and measured target images, respectively. The experimental results demonstrated that the centroid location accuracy of the algorithms used in the experiment was relatively high, reaching the subpixel level. Among them, GTVIWC has the highest location accuracy. The root mean square error (RMSE) of the centroid location for the simulated and measured CCD imaging RRTs reaches 0.0011 pixels and 0.0122 pixels, respectively. The correctness, reliability, and high accuracy of the proposed algorithm are verified.
科研通智能强力驱动
Strongly Powered by AbleSci AI