Vector quantization of contextual information for lossless image compression

Ginesta, X.; Kim, S.P.

doi:10.1109/dcc.1994.305947

Cited by 12 publications

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…So we quantize the parameter ψ into more manageable number of levels. Designing the optimal quantization levels can be done using dynamic programming or other techniques [73,74]. The ultimate goal and hence the cost function is to reduce the total entropy of the error [25].…”

Section: Bias Cancellation Using Context Modeling and Error Feedbackmentioning

confidence: 99%

Adaptive edge-based prediction for lossless image compression

Rahul¹

View full text Add to dashboard Cite

ADAPTIVE EDGE-BASED PREDICTION FOR LOSSLESS IMAGE COMPRESSION by Rahul G. Parthe The increasing use of images in a variety of applications has heightened the need for methods to compress an image. Image compression is needed in efficient management of the two main resources in computing systems; space and communication time. Computer based data compression has been practiced for decades and image compression has been studied since the 70's. Thus compression algorithms are slowly approaching their theoretical limit. Image data modeling in particular, is very difficult due to the higher order correlations that exist in natural continuous images. Though the established compression results are hard to surpass, there are some aspects that can be considered to achieve better performance. The gain might appear small, but still significant, especially when considered from the viewpoint of increasing image size of available images. Prediction based image compression is generally a two-phase process: the first phase is the prediction process/method, and the second is the error modeling and encoding. Prediction is done based on the previously encoded data [15], which gives information about the local characteristics of the image area using some non-adaptive fixed rule set. The simplest technique used as a predictor is the Differential Pulse Code Modulation (DPCM) [16], which assumes the prediction for the current pixel to be the value of the previous pixel in raster scan. The error generated for encoding is thus calculated using the difference Image Adaptivity: It is important to first identify and classify the image areas into edge and non-edge areas, so that each can be treated differently [31]. Once the pixels are designated and separated into two classes, a predictor suitable for the class and local image characteristics can be applied. An approximate model of the activity around the pixel to be predicted can be adaptively generated and used to better tune the predictor for the current

show abstract

Section: Bias Cancellation Using Context Modeling and Error Feedbackmentioning

confidence: 99%

Adaptive edge-based prediction for lossless image compression

Rahul¹

View full text Add to dashboard Cite

show abstract

Lossless compression of continuous-tone images via context selection, quantization, and modeling

1997

IEEE Trans. on Image Process.

207

View full text Add to dashboard Cite

Context modeling is an extensively studied paradigm for lossless compression of continuous-tone images. However, without careful algorithm design, high-order Markovian modeling of continuous-tone images is too expensive in both computational time and space to be practical. Furthermore, the exponential growth of the number of modeling states in the order of a Markov model can quickly lead to the problem of context dilution; that is, an image may not have enough samples for good estimates of conditional probabilities associated with the modeling states. New techniques for context modeling of DPCM errors are introduced that can exploit context-dependent DPCM error structures to the benefit of compression. New algorithmic techniques of forming and quantizing modeling contexts are also developed to alleviate the problem of context dilution and reduce both time and space complexities. By innovative formation, quantization, and use of modeling contexts, the proposed lossless image coder has a highly competitive compression performance and yet remains practical.

show abstract

A derailment-free finite-state vector quantizer with optimized state codebooks

Ginesta

Kim

Proceedings DCC '95 Data Compression Conference

View full text Add to dashboard Cite

Vector quantization of contextual information for lossless image compression

Cited by 12 publications

References 10 publications

Adaptive edge-based prediction for lossless image compression

Adaptive edge-based prediction for lossless image compression

Lossless compression of continuous-tone images via context selection, quantization, and modeling

A derailment-free finite-state vector quantizer with optimized state codebooks

Contact Info

Product

Resources

About