Adaptive perceptual quantizer for high dynamic range video compression

Abstract

Although high dynamic range (HDR) videos are a very fascinating way to represent real-world scenes, they need a huge amount of memory to store and transmit due to the high bit-depth. Thus, it is a major challenge for HDR video coding to efficiently compress them without sacrificing perceptual quality. Perceptual Quantizer (PQ) transfer function provides a solution to this problem, which is adopted as the HEVC Main 10 Profile-based Anchor. However, PQ is not adaptive to HDR contents, thus reducing the coding efficiency. In this paper, we propose adaptive transfer function based on PQ for HDR video compression, called adaptive PQ. Different from PQ which uses a fixed mapping curve from luminance to luma, the proposed transfer function adaptively maps luminance to luma according to HDR contents. Thus, adaptive PQ is able to efficiently utilize possible luma values. Moreover, adaptive PQ achieves better perceptual uniformity in the luminance range than PQ. Experimental results demonstrate that adaptive PQ achieves a significant performance improvement in HDR video coding over PQ in terms of visual quality and bitrate.