在视觉测量中,经常发生椭圆中心点检测错误从而影响图像的匹配问题,甚至导致三维重建任务的失败。针对该问题,设计了1种可纠错的编码点,将信息论中的信道可纠错编码方法用于编码点的设计,基于汉明线性分组码将编码分为信息元和校验元,设计了生成矩阵和校验矩阵,同时在检测出错误时利用伴随矩阵进行纠错,最后进行译码用于三维匹配。实验表明对于只发生1个位置错误的编码可以自动的进行纠错,验证了算法的可行性。
In the measurement of vision, the error often happens when detecting the elliptic point which affects image matching problem, and even cau- ses the failure of 3D registration task. In order to solve this problem, the coding point which can correct the error is designed, and the channel coding er- ror correction method of information theory is used in the design of the coding point. It will be divided into calibration element and information element based on the Hamming linear block code, thus generation matrix and check -up matrix are generated, after that using adjoin matrix to correct the error, then decoding is used for the 3D registration. The result shows that it can correct the error automatically when has an error, which verifies the feasibility of the algorithm.