摘要
在计算机视觉测量系统中,摄像机的标定精度对保证测量结果的可靠性起着至关重要的作用.本文利用数学形态学探讨了图像处理的精度,并论述了豪斯荷尔德矩阵分解法在确保求解数值稳定性中的作用.最后通过对标定块上特征点进行三维重建的精度分析,验证了摄像机的标定精度.
The accuracy of camera's calibration is crucial to the reliability of measurement in the systems of the computer vision. The paper illustrates the precision of image processing by means of mathematical morphology, and discusses Householder matrix decomposing method and its application to keep the numerical stability. Then the accuracy of calibration of cameras have been tested. By analyzing the precision of three-dimensional restructure for the feature points in calibration block.
出处
《新疆大学学报（自然科学版）》
CAS
2004年第1期73-76,共4页
Journal of Xinjiang University(Natural Science Edition)
