A novel approach for analysis of outcrop fracture properties from 3D LiDAR point cloud data

Abstract LiDAR(Light Detection and Ranging) technology is a fully automated,high-precision stereo scanning method.As one of many of its application in geoscience,it enables the rapid acquisition of precise 3D models of natural rock outcrops.However,extracting fracture information directly from 3D point cloud data is challenging,primarily due to the complexity introduced by the reflection of fractures on the surface of the rocks.This paper proposes a point cloud processing approach after the acquisition of point cloud data that transforms 3D point cloud data into 2D images using an innovative gridding method while preserving the maximum amount of fracture information.Noise reduction,Tensor Voting,and the rectangle bounding algorithm are seamlessly integrated into the processing workflow to enhance andquantitatively characterize fracture properties.Statistical data on fracture properties are derived,providing crucial parameters necessary for fractured hydrocarbon reservoir modeling.The feasibility of the novel approach is substantiating data from anoutcrop in the Yangxia Depression,validated by field photographs.By finetuning thresholds and processing parameters,this method can be adapted for use with various other outcrop point cloud datasets,making it an applicable tool for diverse applications.