An Orthogonal 16-point Approximate DCT for Image and Video Compression

TL;DR

This paper introduces a low-complexity, orthogonal, multiplierless 16-point DCT approximation optimized for image and video compression, demonstrating high efficiency and minimal quality loss in practical implementations.

Contribution

It presents a novel 16-point DCT approximation with minimal computational cost, hardware implementation, and integration into HEVC, outperforming existing algorithms.

Findings

Requires only 60 additions for computation

Outperforms classical and state-of-the-art algorithms

Negligible video quality degradation in HEVC

Abstract

A low-complexity orthogonal multiplierless approximation for the 16-point discrete cosine transform (DCT) was introduced. The proposed method was designed to possess a very low computational cost. A fast algorithm based on matrix factorization was proposed requiring only 60~additions. The proposed architecture outperforms classical and state-of-the-art algorithms when assessed as a tool for image and video compression. Digital VLSI hardware implementations were also proposed being physically realized in FPGA technology and implemented in 45 nm up to synthesis and place-route levels. Additionally, the proposed method was embedded into a high efficiency video coding (HEVC) reference software for actual proof-of-concept. Obtained results show negligible video degradation when compared to Chen DCT algorithm in HEVC.