Blind quality scalable video compression algorithm for low bit-rate coding

An efficient way of using Embedded Zero-trees of Wavelet (EZW) architecture for video compression, termed as “Two Threshold EZW”, has been introduced in this paper. This architecture is free from motion compensation and post-compression rate-distortion (PCRD) optimization algorithms. To catch up with the compression obtained by H.265, traditional EZW architecture has been modified for exploiting the temporal/inter-frame dependency. To exploit this, the concept of two thresholds for intra and inter-frame coding has been introduced for a single frame. It is also helpful in coding spatial high-pass and low-pass components independently, which adds more degree of freedom for rate optimization. This architecture makes it easier for achieving blind scalability without using PCRD optimization. To evaluate the algorithm, it has been compared with H.265 at different bit rates. Peak Signal to Noise Ratio (PSNR) and Structural Similarity (SSIM) index are calculated to compare the quality of reconstruction at various low bit rates.