Scanning Order Strategies for Bitplane Image Coding

Abstract

Scanning orders of bitplane image coding engines are commonly envisaged from theoretical or experimental insights and assessed in practice in terms of coding performance. This paper evaluates classic scanning strategies of modern bitplane image codecs using several theoretical-practical mechanisms conceived from rate-distortion theory. The use of these mechanisms allows distinguishing those features of the bitplane coder that are essential from those that are not. This discernment can aid the design of new bitplane coding engines with some special purposes and/or requirements. To emphasize this point, a low-complexity scanning strategy is proposed. Experimental evidence illustrates, assesses, and validates the proposed mechanisms and scanning orders.

General Embedded Quantization for Wavelet-Based Lossy Image Coding

“…The quantizer's complexity is assessed in terms of quantization stages or, equivalently, in terms of coding passes included in the codestream at a target rate. The quantizer's complexity evaluation is relevant since, in general, a codec that codes fewer coding passes spends less computational resources as well [37]. The number of coding passes of the practical GEQ is computed as 2k ′ when k ′ ≤ 3 (recall that each significance coding stage requires 2 coding passes), and as (3k ′ +3)/2 and (3k ′ +2)/2 when k ′ is odd and even, respectively, with k ′ denoting the number of quantization stages coded for that codeblock.…”

“…The wavelet transform [35], [36] is employed by many image coding systems [13], [25], [37] and standards [38], [39] to decorrelate image data. The following analysis and the subsequent practical approach investigate GEQ schemes for the coding of coefficients produced by a wavelet transform.…”

“…JPEG 2000 is chosen due to its widespread use, excellent coding performance and context-based adaptive mechanisms, which are sound for many types of data and coding parameters. As originally formulated, tier-1 codes wavelet coefficients quantized by USDQ performing multiple coding passes per bitplane, which is a common practice of BPC strategies [37]. Herein, tier-1 is modified to allow the coding through the GEQ.…”

“…The derivation ofr k andd k for intervals other than the deadzone is determined similarly as in Equations (4) and (5). An extended description of rate and distortion models similar to those employed in this work can be found in [32], [37], [42].…”

Strategy of Microscopic Parallelism for Bitplane Image Coding

“…In the field of image compression, the most popular techniques to code the codeblock's data are bitplane coding and context-adaptive arithmetic coding [2]. These techniques scan the coefficients within the codeblock sequentially, so that the probability model can adjust the probability estimates of the emitted symbols as more data are coded.…”