Robust Location Algorithm for Planar Target Based on Fuzzy Clustering

Planar target is one of the fundamental accessories in terrestrial laser scanner and plays an important role in multi-view point clouds registration and in calibration of terrestrial laser scanner, among which determining center position of planar target is a key step. At present, there are few studies on locating planar target. Based on introduction of fuzzy c-means clustering algorithm, we proposed a new method to locating center of planar target. The algorithm is divided into three steps: first using fuzzy clustering algorithm to partition planar target data into different types, and then extract the type that we interest and using robust plane fitting algorithm to delete outliers and last obtain target center coordinates. Experimental results show the correctness and robustness of the method.