Fast codebook search algorithm for unconstrained vector quantisation

Chen, C.-Q.; Koh, Soo Ngee; Soon, Ing Yann

doi:10.1049/ip-vis:19981691

Cited by 5 publications

(2 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To remedy this problem, the fast search algorithms partial distance search 5 ͑PDS͒, predictive partial distance search 6 ͑PPDS͒, triangle inequality ͑TIE͒ rule, 7 meandistance-ordered partial codebook search ͑MPS͒ algorithm, 8 double test algorithm, 9 diagonal axes method 10 ͑DAM͒, and many others [11][12][13][14][15][16] have been developed. To remedy this problem, the fast search algorithms partial distance search 5 ͑PDS͒, predictive partial distance search 6 ͑PPDS͒, triangle inequality ͑TIE͒ rule, 7 meandistance-ordered partial codebook search ͑MPS͒ algorithm, 8 double test algorithm, 9 diagonal axes method 10 ͑DAM͒, and many others [11][12][13][14][15][16] have been developed.…”

Section: Introductionmentioning

confidence: 99%

Predictive subcodebook search algorithm for vector quantization

Chang

2000

Opt. Eng

View full text Add to dashboard Cite

A novel variable-rate image-coding scheme called the predictive subcodebook search algorithm (PSS) is proposed. In PSS, the initial searched codeword for each input vector is predicted by its encoded neighbors and the closest codeword in the codebook is searched from a specific range (or subcodebook) that is determined dynamically according to the block type of its encoded neighbors. In addition, the relative addressing technique is employed to encode the searched closest codeword. According to the simulation results, it is shown that PSS not only provides better objective image quality peak SNR (PSNR) but also reduces the storage cost for storing multiple codebooks. Furthermore, PSS requires very little computational cost, compared to the full search algorithm for the vector quantization (VQ) scheme. Therefore, it is concluded that PSS indeed provides a good means for variable-rate image coding. Downloaded From: http://opticalengineering.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

show abstract

Section: Introductionmentioning

confidence: 99%

Predictive subcodebook search algorithm for vector quantization

Chang

2000

Opt. Eng

View full text Add to dashboard Cite

show abstract

“…First, derivation of a representative set ofpattems to form a compact codebook has been an elusive goal of VQ research for several decades [14][15][16]. Second, given a representative codebook of Q exemplars and a source image of N pixels subdivided into a collection of K-pixel neighborhoods, the codebook search overhead involved in matching source patterns to codebook patterns approaches a minimum of NQ comparisons for Boolean images and 2NQ additions for greyscale imagery [17]. If Q and N are large (e.g., 1M pixel or larger) and K is of moderate size (e.g., 16 to 100 pixels), then such overhead can be prohibitive for real-time compression of image sequences at video rates.…”

Section: Introductionmentioning

confidence: 99%

<title>Performance analysis of tabular nearest-neighbor encoding for joint image compression and ATR: I. Background and theory</title>

1999

View full text Add to dashboard Cite

Vector quantization (VQ), a well-established image compression technique, is attractive due to its mathematically simple compression formulation and efficient decompression implementation. Unfortunately, VQ typically suffers from high cost of codebook computation, as well as relatively high codebook search overhead during compression. Such costs can be prohibitive when VQ is employed in real-time surveillance applications, especially with large-format images and moderate to high compression ratios.In response to this situation, Frontier Technology, Inc. has developed an efficient codebook search technique called Tabular Nearest Neighbor Encoding (ThE). This approach significantly reduces the codebook search cost for VQ-and VQ-like compression algorithms, and can increase pattern-matching accuracy and efficiency in template-based automated target recognition (ATR) applications. In particular, TNE's VQ-like codebook can represent a database of target patterns for ATR purposes, which tends to yield an increase in pattern matching efficiency due to the processing of fewer (encoded) data.In this series of two papers [37], a high-level overview of the TNE algorithm is presented. The performance of TNE is analyzed using training images having different size, statistical properties, and noise level than the source image. ThE is compared with several published algorithms such as visual pattern image coding (VPIC), JPEG, and EBLAST. The latter is a relatively new, high-compression image transform that has compression ratio CR 200:1 that can be consistently achieved with low MSE. Analysis focuses on the ability of ThE to provide low to moderate compression ratios at high computational efficiency on small-to large-format text and surveillance images.

show abstract