Lossy Dictionaries

Lossy Dictionaries

Two-Stage Algorithm for Data Compression

This paper proposes two stage algorithm that carries advantages of PDLZW and Arithmetic coding and compares its performance with deflate which is a well-known two-stage algorithm that combines the features of LZ77 and Huffman Coding. The PDLZW is designed by partitioning the dictionary into several dictionaries of different address spaces and sizes. With the hierarchical parallel dictionary set...

Improving SOM Vector Quantization for Image Compression with Walsh-Hadamard Transform

The bandwidth reduction or storage lowering in digital image transmission confers to the image compression a key role. In this paper, we propose a new approach for lossy image compression: the source image is vector quantized by applying Self-Organizing Map (SOM) with several dictionaries. Each dictionary is originally designed based on the feature vectors resulted after applying the Walsh-Hada...

Image Compression by Parameterized-Model Coding of Wavelet Packet Near-Best Bases

Top-down tree search algorithms with non-additive information cost comparisons as decision criteria have recently been proposed by Taswell for the selection of near-best bases in wavelet packet transforms. Advantages of top-down non-additive near-best bases include faster computation speed, smaller memory requirement, and extensibility to biorthogonal wavelets in addition to orthogonal wavelets...

An efficient first-order method for l1 compression of images

We consider the problem of lossy compression of images using sparse representations from overcomplete dictionaries. This problem is in principle easy to solve using standard algorithms for convex programming, but often the large dimensions render such an approach intractable. We present a highly efficient method based on recently developed first-order methods, which enables us to compute sparse...