Abstract
In this document, we present new techniques for near-lossless and lossy compression of SAR imagery saved in PNG and binary formats of magnitude and phase data based on the application of transforms, dimensionality reduction methods, and lossless compression. In particular, we discuss the use of blockwise integer to integer transforms, subsequent application of a dimensionality reduction method, and Burrows-Wheeler based lossless compression for the PNG data and the use of high correlation based modeling of sorted transform coefficients for the raw floating point magnitude and phase data. The gains exhibited are substantial over the application of different lossless methods directly on the data and competitive with existing lossy approaches. The methods presented are effective for large scale processing of similar data formats as they are heavily based on techniques which scale well on parallel architectures.
Highlights
We present new techniques for near-lossless and lossy compression of Synthetic-aperture radar (SAR) imagery saved in PNG and binary formats of magnitude and phase data based on the application of transforms, dimensionality reduction methods, and lossless compression
We provide some details on different SAR data formats, which are common to various remote sensing type data, discuss blocking and transform application, which is an essential parallelizable step for the redundancy reduction direction, and present the proposed algorithms for integer and floating point data which combine redundancy reduction and compression techniques
In the case of the former, a Burrows-Wheeler local similarity transform (BWT) is used to reversibly transform the input into a better compressible form based on string sorting, followed by a variant of the move to front (MTF) transform and run length encoding (RLE) for rearranging frequently appearing symbols to the front and compressing sequences of identical data and entropy coding (EC), such as Huffman or arithmetic coding, for more efficiently representing the input data
Summary
SAR data are often presented in complex form, with real and imaginary parts or equivalently in polar format with magnitude and phase information [2]. Effective compression of complex SAR data is often necessary for storage and transmission applications, yet often challenging due to the need for high reconstruction accuracy, in applications such as interferometry. The techniques applied here can vary based on the available format of the SAR data, either encoded as PNG, or in raw format with magnitude and phase components. In both cases, it is interesting to exploit redundancy in the data via lossy compression, prior to the application of lossless compression techniques. We propose two different methods to do this, one tailored for sets of integer data and one for floating point data
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
