Font Size: a A A

Research On Compression Algorithm For 3D Point Cloud Model

Posted on:2021-05-25Degree:MasterType:Thesis
Country:ChinaCandidate:M Y ZhangFull Text:PDF
GTID:2428330611970881Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
With the development of multimedia communication and three-dimensional imaging technology,three-dimensional point cloud models are widely used.The 3D point cloud model is composed of many points,and each point is composed of position,color,and even normal vector information.Its huge data volume makes it very limited in storage and real-time transmission,so the compression of the 3D point cloud model is extremely necessary.This article studies the 3D point cloud compression algorithm,the main content and results are as follows:(1)A new slicing point-eliminating algorithm is proposed,which is a hierarchical point cloud reduction algorithm.It divides each model into several sub-slices,uses angle limit difference and chord height limit difference to remove non-key points,solves the problem of missing key points caused by random sampling method,uniform grid method,normal vector method.Compared with the normal vector angle method,the slicing point-eliminating improves the fidelity of the algorithm,and the P-PSNR value and P-PSNR_y can reach about 35dB,under the same reduction rate.(2)A method based on IAAC hybrid compression algorithm are proposed.Method one combines the IAAC method with JPEG 2000,method two combines IAAC with EDP.Both of the above are to project the three-dimensional point cloud model data to the two-dimensional plane,and use the above-mentioned hybrid method to compress the color data of the point cloud.Aiming at the position and color information of the point cloud,method three combines IAAC and Haar DWT to perform wavelet transform and,and uses the IAAC method for the transformed coefficients.A strategie based on IAAC hybrid compression algorithm are proposed.Simulation result show:Among several strategies based on the IAAC hybrid compression algorithm:the compression ratio of Strategy 1 is improved compared to JPEG 2000,the PSNR and P-PSNR_y values are around 30 dB,and the PSNR and P-PSNR_y of Strategy 2 are above 30 dB,respectively;Strategy 3 compared with the existing wavelet transform method,P-PSNR is slightly lower,but the compression rate is improved 10%.
Keywords/Search Tags:point cloud reduction, point cloud coding, IAAC, wavelet transform, quality evaluation
PDF Full Text Request
Related items