摘要
以摄像机标定作为研究背景,利用黑白棋盘网格特点,结合平面射影变换及最小二乘法,开发了人机交互棋盘网格直线检测系统。该系统只需在图像中手动选取四个角点,然后系统利用模板点和图像点的对应关系求出平面间的单应矩阵,并自动检测出图像中的直线特征。采用距离、倾斜度、旋转度以及尺寸等参数对黑白棋盘网格图像进行实验,验证了方法的可行性。
For camera calibration,an interaction grid line detection system is discussed,by using the black and white checkerboard grid images,plane projective transformation and the least square method.With the relationship between the templates and corners,only four corners of the image are chosen manually to get the homograph to detect the line characters of the image.Some indexes such as distances,slope,rotation and size are used to test the black and white checkerboard grid images,which verifies the feasibility of the method.
作者
李晓芳
LI Xiaofang(College of Computer and Information Engineering, Shanxi Technology and Business College, Taiyuan 030000, China)
出处
《长春工业大学学报》
CAS
2020年第4期387-394,共8页
Journal of Changchun University of Technology
基金
山西省教育科学“十三五”规划课题(GH-19156)。
关键词
直线检测
HOUGH变换
黑白棋盘网格
摄像机标定
最小二乘法
line detection
Hough transform
black and white checkerboard grid
camera calibration
the least square method