亚像素渲染
卡钳
计算机科学
灰度
算法
重心
噪音(视频)
激光器
像素
人工智能
直线(几何图形)
计算机视觉
点(几何)
光学
数学
图像(数学)
几何学
物理
经济
管理
作者
Wenqing Hou,Huicheng Jing,Ao Guo,Yaoqi Chen,Dashun Song
标识
DOI:10.1016/j.optlastec.2023.109743
摘要
The center line extraction of line structured light strip is very important for 3D measurement. A new sub-pixel extraction method of laser center line is proposed. Firstly, the RC-YOLOv5s algorithm is used to extract the effective area of the light strip, which accelerates the extraction speed of the laser light strip and avoids the influence of background noise. An improved Gaussian curve fitting algorithm is used to initially extract the subpixel centerline of the light strip, and the normal direction is calculated based on the initial extracted point. Improved the grayscale center of gravity method based on normal vectors, which is used to calculate the subpixel centers of light strips and achieve high-precision 3D reconstruction of brake calipers. The experimental results show that the proposed algorithm has a significant improvement in the accuracy of light strip extraction, with a lower runtime than the Steger algorithm, and a significant improvement in the effectiveness of 3D reconstruction.
科研通智能强力驱动
Strongly Powered by AbleSci AI