Back to articles
Articles
Volume: 31 | Article ID: art00013
Abstract

Our goal is to develop methods for lossless encoding of automotive lidar sensor data with very low computational complexity and high compression ratio. In this paper, we propose a solution that is based on organizing and packing lidar data into a 2-D image array and subsequently using existing image compression methods. This approach leverages image compression technology that has been developed and proven over many years of R&D, standardization, and wide deployment. In our approach, the X,Y,Z coordinates of lidar scan points are quantized, packed into one or more 2-D images, and subsequently compressed by an image codec. In addition, lidar scan points are re-ordered to optimize spatial prediction and compression efficiency. We have obtained initial results on automotive lidar data scans using several compression engines. Results using PNG and JPEG-LS and using very simple packing techniques show significant compression gains over traditional lidar data coding methods.

Subject Areas :
Views 102
Downloads 25
 articleview.views 102
 articleview.downloads 25
  Cite this article 

Peter van Beek, "Image-based compression of LiDAR sensor datain Proc. IS&T Int’l. Symp. on Electronic Imaging: Autonomous Vehicles and Machines Conference,  2019,  pp 43-1 - 43-7,  https://doi.org/10.2352/ISSN.2470-1173.2019.15.AVM-043

 Copy citation
  Copyright statement 
Copyright © Society for Imaging Science and Technology 2019
72010604
Electronic Imaging
2470-1173
Society for Imaging Science and Technology
7003 Kilworth Lane, Springfield, VA 22151 USA