Band ordering in lossless compression of multispectral images

Abstract

(c) 2011 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, including reprinting/ republishing this material for advertising or promotional purposes, creating new collective works for resale or redistribution to servers or lists, or reuse of any copyrighted components of this work in other works. Abstract:In this paper, we consider a model of lossless image compression in which each band of a multispectral image is coded using a prediction function involving values from a previously coded band of the compression, and examine how the ordering of the bands affects the achievable compression.We present an efficient algorithm for computing the optimal band ordering for a multispectral image. This algorithm has time complexity O(n 2 ) for an n-band image, while the naive algorithm takes time (n!). A slight variant of the optimal ordering problem that is motivated by some practical concerns is shown to be NP-hard, and hence, computationally infeasible, in all cases except for the most trivial possibility.In addition, we report on our experimental findings using the algorithms designed in this paper applied to real multispectral satellite data. The results show that the techniques described here hold great promise for application to real-world compression needs.Index Terms-Compression, lossless compression, image compression, multispectral images, satellite data, NPcompleteness. Article:1 INTRODUCTION MULTISPECTRAL satellite images require enormous amounts of space, and with NASA's project EOS (the Earth Observing System), data will be generated at an unprecedented rate. The estimates are that over a terabyte (10 12 bytes) of data will be generated every day by the EOS satellites, most of it multispectral image data. Largely due to this fact, a lot of attention has recently been focused on compression of multispectral images [5]. However, most of the compression methods that exploit spectral as well as spatial redundancy have been lossy compression algorithms, and for archival storage and for certain applications it is important to use lossless compression in order to preserve all of the data that is collected. One notable exception to this is the work of Roger and Cavenor [4] who extensively study various prediction and coding methods used for lossless compression of AVIRIS data. Table 1 lists some current, widely used multispectral sources, with their acronyms, full names, and basic properties-it is data from these sensors that we used in this study.In this paper, we study lossless compression of multispectral images. Spectral redundancy is extracted by coding each band of the multispectral image by making use of a second -prediction band.‖ In much the same way that standard single-image lossless compression is separated into the two separate components of prediction and coding, we divide the lossless compression of multispectral images into three components: band ordering, prediction, and coding. The new stage, band ordering, refers to selecting a permutation...

Compression of Spectral Images

“…In (Tate, 1997) the actual compression ratio for the data from a single AVIRIS image was 3.53. In (Memon, et al, 1994) the entropy of the residual of Cuprite image ranged from 5.48 to 5.61 bits/sample.…”

“…Also other lossless coding methods exist, e.g. they are based on band ordering (Tate, 1997;Toivanen et al, 2005) or spectral and spatial noncausal prediction (Memon et al, 1994). Thus, in the lossless compression of spectral images better results are achieved through the transform coding and especially, with the predictive coding.…”

A Novel Adaptive Compression Technique for Dealing with Corrupt Bands and High Levels of Band Correlations in Hyperspectral Images Based on Binary Hybrid GA-PSO for Big Data Compression

“…Due to the importance of generating highly accurate information about the atmosphere, clouds, and surface parameters provided by the hyperspectral sensors, lossy compression techniques are not acceptable in this case [2]. The economics of transmission and mass-storage of the large volumes of data accumulated by these sensors demonstrate that efficient compression is very important in this technology [3]. There are a few well-known methods for lossless compression, such as JPEG standards.…”

Lossless Predictive Compression of Hyperspectral Images

“…A significant improvement in coding performance of predictive coding algorithms used with hyperspectral images can be obtained if the bands are reordered to enhance the band-to-band predictability [10]. The motivation behind this idea is that the reflection intensity of the frequency components by materials in the earth generally are not monotonically increasing or decreasing.…”