摘要
基于椭圆曲线拟合和曲率滤波,提出同心圆投影中心定位方法。采用单应矩阵变换和坐标平移法生成了大量的理想测试图像,并与角点检测法的测量精度进行了比较,结果表明,文中方法的测量精度在有无噪声情况下均高于角点检测方法。当信噪比介于35dB和25dB之间时,角点检测法的测量误差突然急剧增大,文中方法的测量误差变化仍比较平缓,测量误差小于0.15pixel。
A location method for concentric circles'projection center based on ellipses’curve fitting and curvature filter is put forward . The test images are generated with homography transform and coordinate shift ,and the method is compared with the corner detection .The results show that the accuracy of the method is better than that of corner detection no matter how the images have noise or not .When images’ S/N ratio is between 25 - 35 dB , the error of the corner detection increases sharply w hile that of concentric circle method is only up to 0 .15 pixel .
出处
《长春工业大学学报》
CAS
2014年第5期500-505,共6页
Journal of Changchun University of Technology
基金
国家重大科研装备研制基金资助项目(ZBYZ2008-1)
关键词
机器视觉
图像处理
同心圆定标
椭圆拟合
machine vision
image processings
concentric circle calibration
ellipse fitness