Motion compensated compression for event-based cameras
Loading...
Date
2022
Authors
Bairagi, Arnob Kumar
University of Lethbridge. Faculty of Arts and Science
Journal Title
Journal ISSN
Volume Title
Publisher
Lethbridge, Alta. : University of Lethbridge, Dept. of Mathematics and Computer Science
Abstract
Event cameras detect significant changes at each pixel asynchronously and report these events in real-time. A changing scene can generate many events in a short time. Efficient storage and transmission are necessary for further processing of this event data. Inspired by this necessity, we propose a lossless Motion Compensated Compression algorithm based on Optical Flow (MCCOF) for event cameras. We analyzed our proposed algorithm performance compared with the lossless spike coding algorithm. We found that our MCCOF algorithm achieves a higher compression ratio on most datasets compared to the spike coding algorithm. Using a real-time event-based optical flow algorithm for motion compensation, our algorithm does not significantly increase the computational time for compression and decompression.
Description
Keywords
event-based cameras , motion compensated compression , event data processing , lossless compression algorithm , optical flow algorithm , log-luminance , Image stabilization , Data compression (Computer Science) , Algorithms , Image processing , Optical data processing , Image reconstruction , Dissertations , Academic