Robot Vehicle Video Image Compression. Phase 1
Abstract
This Phase I SBIR study analyzed and simulated in software a new approach to image data compression for remote vehicle driving. Results show feasibility for 1,000-to-1 compression rates for full color, wide field of view, real time imagery. The approach is based on a three stage process. The first stage consists of reducing resolution of peripheral imagery to match the drop in human visual resolution away from the center of view. This is accomplished by resampling video camera output in real time. The second stage consists of coding color and contrast edges in separate channels. The third stage consists of applying a simple, efficient data compression technique. A hybrid DCT/DPCM data compression method proved to be the most suitable. All three stages of the compression process and reconstruction of imagery can be implemented in real time on equipment which can be compactly mounted in remote vehicles. This is a fundamentally new and valuable attack on the problem of remote driving, which has been with us for decades. The solution is critical for deployment of Army remote systems.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1989
- Accession Number
- ADA210974
Entities
People
- Carl F. Weiman