Home // CONTENT 2018, The Tenth International Conference on Creative Content Technologies // View article


An Efficient HDR Video Compression Scheme based on a Modified Lab Color Space

Authors:
Maryam Azimi
Mahsa T. Pourazad
Panos Nasiopoulos

Keywords: High Dynamic Range (HDR); color encoding; perceptual quantizer; color difference, CIELAB

Abstract:
With recent developments in both High Dynamic Range (HDR) capturing and display technologies, consumer distribution of HDR videos is now possible. The current distribution pipeline, however, is based on Standard Dynamic Range (SDR) signal characteristics. To accommodate HDR in the current pipeline infrastructure, some adjustments are necessary. One element of the pipeline that needs to be addressed is compression scheme. Current compression standards, such as HEVC, rely on YCbCr as their color encoding. However, this color encoding cannot represent the HDR signal without introducing visual artifacts. Two approaches exist for compensating for these visual errors: re-adjustment of the HDR signal in YCbCr as a pre-processing step to compression, or employing a color encoding that better meets the HDR signal requirements. In this paper, we propose to use the perceptually uniform CIELAB color space for HDR video color encoding following the latter approach. To make CIELAB suitable for HDR video color encoding, we change the transfer function and the scaling of the a* and b* channels. The performance of this approach is compared to YCbCr color encoding in our study via compressing these signals based on High Efficiency Video Coding (HEVC) standard and comparing the bitrate of these signals at the same quality level. An average bit-rate saving of 12.9% in terms of Overall Signal to Noise Ratio (OSNR), and 41.3% in terms of DE100 are reported for our proposed color encoding scheme compared to the conventional YCbCr. A negligible average loss of 0.6% is reported in terms of perceptually transformed Peak Signal to Noise Ratio (tPSNR).

Pages: 18 to 23

Copyright: Copyright (c) IARIA, 2018

Publication date: February 18, 2018

Published in: conference

ISSN: 2308-4162

ISBN: 978-1-61208-611-8

Location: Barcelona, Spain

Dates: from February 18, 2018 to February 22, 2018