Survey of Image Compression Techniques with Applicability to SAR Imagery.

摘要

This paper is intended as a primer on image compression techniques as they apply to compressing Synthetic Aperture Radar (SAR) imagery. Particular attention will be focused on two techniques which are thought to be particularly suitable to this particular application because of their simplicity and speed, namely Vector Quantization (VQ) and Block Truncation Coding (BTC). More well-known techniques such as Differential Pulse Code Modulation (DPCM) and Discrete Cosine Transform (DCT) will also be discussed. Recommendations for experiments on SAR imagery are made. SAR imagery differs from conventional visual band photographic or electro-optical imagery in two significant respects: first by having a large dynamic range (as great as 96 dB), and second, by the presence of speckle noise, the sort of noise seen in optical holography.