Abstract: This paper presents a method of similarity analysis algorithm of the three -dimensional point cloud,which is based on eigenvector of the subspace. First of all, the three-dimensional point cloud data of two objects were obtained and positions of them were standardized. And then, the two three - dimensional point clouds were divided into several subspace by using the minimal spatial segmentation algorithm. Thirdly, the eigenvector of subspace were calculated, which should be divided into two steps: the first step was to calculate distance and angle from the centroid to the subspace surface, the next step was to compute the new eigenvector on the basis of vector space, which was composed of the distance and angle in step one. This research method took the advantage of small data in quantity and high precision in calculation because the eigenvector of subspace, which can describe the three -dimensional characteristics as the basis of similarity measure. The experiment shows that the algorithm can quantitatively analyze the similarity of two three-dimensional objects.