Basic concepts of knowledge-based image understanding
KES-AMSTA'08 Proceedings of the 2nd KES International conference on Agent and multi-agent systems: technologies and applications
On the empirical rate-distortion performance of compressive sensing
ICIP'09 Proceedings of the 16th IEEE international conference on Image processing
Matching pursuits with time-frequency dictionaries
IEEE Transactions on Signal Processing
IEEE Transactions on Information Theory
| Hi-index | 0.00 |
We verified compression efficiency of the procedures based on compressive sensing (CS) inspiration. Medical imaging was concerned as challenging area of possible applications. Irreversible image compression algorithm was integrated with source data measurements according to CS rules. Two kinds of measurements as inner products against adjusted atoms were used: regular cosines and pseudo-random noiselets. Image coarse representation was approximated from linear cosine measurements while important details were estimated basing on fixed noiselet measurements. Simulated sensor system projected the image onto a set of separable 2-D basis functions to measure the corresponding expansion coefficients. Such procedure was optimized and augmented to construct integrated method of image sensing, compression and data processing. We proposed algorithm of selected measurements with uniformly quantized coefficients formed and encoded with necessary side information. Universal PAQ8 archiver was used to complete compression procedure. Experimentally verified compression schemes showed possible compression improvement by designed procedure in comparison to reference JPEG and JPEG2000 encoders.