Linear Feature Extraction And Deformation Analysis Of Unstable Rock Mass Surface

Ground based 3D terrestrial laser scanning is widely used in deformation monitoring for its characteristics of acquiring point cloud data of observed surfaces in non-contacting, fast, high precision, high density way. Usually, we utilize the change of the feature of the target surface to monitor the deformation, but the data obtained by the three demensional terrestrial laser scanner is discrete, with no attribute information and topology relation. It is useful to apply the two-dimensional gray image feature extraction method to the point cloud data feature extraction, namely resampling the point cloud data, generating grid image and then using image processing algorithm for feature extraction. Thus, we can analyze the deformation by the extracted feature. The main work of this paper is as follows:(1) Take the three phase of the Chongqing jiguan ridge point cloud data as the data source of this paper. Firstly, register the three phase data roughly, then use ICP algorithm for more accurate registration.(2) Interpolate the discrete data of different phase into grid data by Kriging method, then change the grid data into depth image, extract linear feature in same position by the digital image processing technique.(3) Based on the extracted linear feature, we can compare the variation of point in linear feature of different phase data to analyze the state of the unstable rock mass, thus, we get the deformation change of the unstable rock mass surface.Many research focus on the deformation monitoring methods. This paper tries to explore that project by analyzing the change of the linear feature, which provides view for unstable rock mass deformation monitoring.