Joan Serra-Sagristà

Extending the CCSDS Recommendation for Image Data Compression for Remote Sensing Scenarios

This paper presents prominent extensions that have been proposed for the Consultative Committee for Space Data Systems Recommendation for Image Data Compression (CCSDS-122-B-1). Thanks to the proposed extensions, the Recommendation gains several important featured advantages: It allows any number of spatial wavelet decomposition levels; it provides scalability by quality, position, resolution, and component; and it supports multi-/hyper-/ultraspectral data coding, allowing a spectral decorrelation if requested. As a consequence, compression performance is notably improved with respect to the Recommendation for a large variety of remote sensing images, both monoband and multi-/hyper-/ultraspectral images. Reported results for hyperspectral data suggest that our proposal is competitive with the JPEG2000 standard.

A Tutorial on Image Compression for Optical Space Imaging Systems

Public policies and private initiatives share the will to explore outer space and to monitor the Earth from space sensors. Recent years have seen an increased number of space missions, while the sensors on board aircrafts or spacecrafts have also significantly improved their acquisition capabilities. Given this huge volume of remote sensing data and the detailed characteristics of the acquired images, a data compression process is in order to allow as large a transmission rate as possible.

Cost and Scalability Improvements to the Karhunen–Loêve Transform for Remote-Sensing Image Coding

The Karhunen-Loêve transform (KLT) is widely used in hyperspectral image compression because of its high spectral decorrelation properties. However, its use entails a very high computational cost. To overcome this computational cost and to increase its scalability, in this paper, we introduce a multilevel clustering approach for the KLT. As the set of different multilevel clustering structures is very large, a two-stage process is used to carefully pick the best members for each specific situation. First, several candidate structures are generated through local search and eigenthresholding methods, and then, candidates are further screened to select the best clustering configuration. Two multilevel clustering combinations are proposed for hyperspectral image compression: one with the coding performance of the KLT but with much lower computational requirements and increased scalability and another one that outperforms a lossy wavelet transform, as spectral decorrelator, in quality, cost, and scalability. Extensive experimental validation is performed, with images from both the AVIRIS and Hyperion sets, and with JPEG2000, 3D-TCE, and CCSDS-Image Data Compression recommendation as image coders. Experiments also include classification-based results produced by k-means clustering and Reed-Xiaoli anomaly detection.

JPEG2000 Quality Scalability Without Quality Layers

Quality scalability is a fundamental feature of JPEG2000, achieved through the use of quality layers that are optimally formed in the encoder by rate-distortion optimization techniques. Two points, related with the practical use of quality layers, may need to be addressed when dealing with JPEG2000 code-streams: 1) the lack of quality scalability of code-streams containing a single or few quality layers and 2) the rate-distortion optimality of windows of interest transmission. Addressing these two points, this paper proposes a mechanism that, without using quality layers, provides competitive quality scalability to code-streams. Its main key-feature is a novel characterization of the code-blocks rate-distortion contribution that does not use distortion measures based on the original image, or related with the encoding process. Evaluations against the common use of quality layers, and against a theoretical optimal coding performance when decoding windows of interest or when decoding the complete image area, suggest that the proposed method achieves close to optimal results.

Low Complexity JPEG2000 Rate Control Through Reverse Subband Scanning Order and Coding Passes Concatenation

This letter introduces a new rate control method devised to provide quality scalability to JPEG2000 codestreams containing a single or few quality layers. It is based on a Reverse subband scanning Order and a coding passes Concatenation (ROC) that does not use distortion measures based on the original image. The proposed ROC method allows a flexible rate control when the image has already been encoded, using negligible computational resources and obtaining the same efficiency as when using quality layers. Besides, the proposed ROC can be used in the encoding process to reduce the coder complexity, avoiding to encode unnecessary coding passes and achieving a competitive performance in terms of mean-square error (MSE)

Clustered Reversible-KLT for Progressive Lossy-to-Lossless 3d Image Coding

The RKLT is a lossless approximation to the KLT, and has been recently employed for progressive lossy-to-lossless coding of hyperspectral images. Both yield very good coding performance results, but at a high computational price. In this paper we investigate two RKLT clustering approaches to lessen the computational complexity problem: a normal clustering approach, which still yields good performance; and a multi-level clustering approach, which has almost no quality penalty as compared to the original RKLT. Analysis of rate-distortion evolution and of lossless compression ratio is provided. The proposed approaches supply additional benefits, such as spectral scalability, and a decrease of the side information needed to invert the transform. Furthermore,since with a clustering approach, SERM factorization coefficients are bounded to a finite range, the proposed methods allow coding of large three dimensional images within JPEG2000.

Efficient rate control for JPEG2000 coder and decoder

The JPEG2000 standard does not specify how to perform the rate control strategy, needed either to achieve a target bitrate, or to construct quality layers. In this paper, an efficient rate control algorithm is described. It is based on the interleaving of coding passes and it has a low computational complexity. The proposed algorithm encodes only the coding passes included in the final codestream, and it fully avoids the need of any post compression rate distortion stage. Extensive experimental results show that the encoding performance of our method is competitive and similar to the optimal strategy. In addition, this proposal allows the extraction of a target bitrate from a codestream without the need of either knowing the original image, or of decoding any part of the codestream. The performance of this kind of extraction is equivalent to that obtained when decompressing a codestream organized in quality layers.

Regression Wavelet Analysis for Lossless Coding of Remote-Sensing Data

A novel wavelet-based scheme to increase coefficient independence in hyperspectral images is introduced for lossless coding. The proposed regression wavelet analysis (RWA) uses multivariate regression to exploit the relationships among wavelet-transformed components. It builds on our previous nonlinear schemes that estimate each coefficient from neighbor coefficients. Specifically, RWA performs a pyramidal estimation in the wavelet domain, thus reducing the statistical relations in the residuals and the energy of the representation compared to existing wavelet-based schemes. We propose three regression models to address the issues concerning estimation accuracy, component scalability, and computational complexity. Other suitable regression models could be devised for other goals. RWA is invertible, it allows a reversible integer implementation, and it does not expand the dynamic range. Experimental results over a wide range of sensors, such as AVIRIS, Hyperion, and Infrared Atmospheric Sounding Interferometer, suggest that RWA outperforms not only principal component analysis and wavelets but also the best and most recent coding standard in remote sensing, CCSDS-123.

Divide-and-Conquer Strategies for Hyperspectral Image Processing: A Review of Their Benefits and Advantages

In the field of geophysics, huge volumes of information often need to be processed with complex and time-consuming algorithms to better understand the nature of the data at hand. A particularly useful instrument within a geophysicists toolbox is a set of decorrelating transforms. Such transforms play a key role in the acquisition and processing of satellite-gathered information, and notably in the processing of hyperspectral images. Satellite images have a substantial amount of redundancy that not only renders the true nature of certain events less perceivable to geophysicists but also poses an issue to satellite makers, who have to exploit this data redundancy in the design of compression algorithms due to the constraints of down-link channels. This issue is magnified for hyperspectral imaging sensors, which capture hundreds of visual representations of a given targeteach representation (called a component or a band) for a small range of the light spectrum. Although seldom alone, decorrelation transforms are often used to alleviate this situation by changing the original data space into a representation where redundancy is decreased and valuable information is more apparent.

Remote Sensing Data Compression

The interest in remote sensing images is growing at an enormous pace in the last years. However, transmission and storage of remote sensing images pose a special challenge, and multiple efficient image compression systems have appeared. This chapter contributes an overview of several techniques for image coding systems, focusing on lossy approaches.