Study On Point Cloud Processing Method And Its Application In Cultural Relic Virtual Recovery

The Terracotta Warriors as the eighth wonder of world,is an important carrier for the scholars study ancient Chinese political,economic,cultural and any other information. While the excavated terracotta figures have been damaged,the data that digitized collected from the broken Terracotta Warriors fragments by means of reverse engineering usually contain a lot of noise,and these point cloud data have a huge amount, that will make the follow-up virtual split work more complex.For the demand of cultural relic virtual recovery,in this paper,we research the optimized method for the point cloud model,the research works as follow:1 Propose a cultural relic point cloud model denoising and smoothing method based on the LOP operator.Point cloud model denoising uses the weighted LOP operator,WLOP operator,solve the problem that the LOP operator process the edge of model unwell in point cloud denoising process.In point cloud model smoothing processing,to improve LOP operator by Scattered point cloud Preserving Interpolation method, effectively avoiding the local volume of point cloud model shrinkage problem in smoothing process, improve the quality of point cloud model.2 Propose a combined point cloud simplification algorithm based on particle simulation and adaptive curvature sampling, The algorithm for the shortage that the particle simulation simplify may appear the blank area and the adaptive curvature sampling easily lead to the loss of point cloud model details,using a combined algorithm of the two methods to simplify the point cloud model,solve the problem of inefficiency and poor simplify effect that traditional algorithm appeared. While streamline point cloud size,effectively retain a significant geometric features of the original model at the same time, to provide a reliable source of data for subsequent processing.3 Design and implement an optimization system based on the Terracotta point cloud model, the system has automatic point cloud denoising, scattered homogenization and point cloud simplified that features keeped and so on. To put the above research methods in use for the optimization process of the cultural relic model data, retaining the detail features of the cultural relic model better. It laid the data foundation for the subsequent heritage splicing.