Share Email Print

Spie Press Book

Optical Satellite Signal Processing and Enhancement
Author(s): Shen-En Qian
Format Member Price Non-Member Price

Book Description

This book gives a global review of optical satellite signal processing theories, algorithms, and system implementations. Consisting of 14 chapters, it describes a variety of spaceborne optical sensors and their performance measurements. The author shares with readers his firsthand experience and research outcomes in developing novel signal processing solutions to handle satellite data and to enhance the satellite sensor performance and the know-how for optical satellite data generation, onboard data compression, and its implementation strategy, data formatting, channel coding, calibration, and various signal processing technologies. Written with both postgraduate students and advanced professionals in mind, this useful monograph addresses the current lack of literature about satellite-based optical imaging and presents a rigorous treatment of the broad array of technologies involved in the field.

Book Details

Date Published: 8 November 2013
Pages: 554
ISBN: 9780819493286
Volume: PM230

Table of Contents
SHOW Table of Contents | HIDE Table of Contents

Table of Contents

List of Terms and Acronyms

1 Spaceborne Optical Sensors
1.1 Introduction
1.2 Optical Satellite Sensors and Their Types
1.3 Panchromatic Sensors
1.4 Multispectral Sensors
     1.4.1 Landsat MSS, TM, and ETM+
     1.4.2 SPOT's HRV, HRVIR, and HRG
     1.4.3 Other multispectral sensors
1.5 Hyperspectral Sensors
     1.5.1 What is a hyperspectral sensor?
     1.5.2 Operating principle of a hyperspectral sensor
     1.5.3 Types of hyperspectral sensors
 Dispersing-element-based sensors
 Optical-filter-based sensors
 Electronically tunable-filter-based sensors
     1.5.4 Hyperspectral sensor operating modes
 Whisk-broom mode
 Push-broom mode
     1.5.5 Spaceborne hyperspectral sensors
1.6 Imaging Fourier Transform Spectrometer Sensor
     1.6.1 Description
     1.6.2 Types of FTS sensors, and operational concept
     1.6.3 Spaceborne IFTS
1.7 Lidar Sensor
     1.7.1 Definition and description
     1.7.2 Lidar In-space Technology Experiment
     1.7.3 Shuttle Laser Altimeter
     1.7.4 Mars Orbiter Laser Altimeter
     1.7.5 Geoscience Laser Altimeter System
     1.7.6 Cloud-Aerosol Lidar with Orthogonal Polarization
     1.7.7 Atmospheric Laser Doppler Lidar Instrument
     1.7.8 Mercury Laser Altimeter
     1.7.9 Lunar Orbiter Laser Altimeter
     1.7.10 Next-generation, high-resolution swath-mapping lidar

2 Satellite Data Generation and Product Levels
2.1 Space Data and Information System
2.2 EOS Data and Information System
     2.2.1 Spacecraft command-and-control center
     2.2.2 Data capture and Level-0 processing
     2.2.3 Product generation
     2.2.4 Data archive, management, and distribution
     2.2.5 Locating and accessing data products of interest
2.3 EOS Data Product Levels
2.4 Planetary Data System and Product
     2.4.1 Standard data products
     2.4.2 Engineering and other ancillary data products
     2.4.3 Dataset documentation
2.5 Planetary Data Product Levels
2.6 Example of EOS Data Product Levels
     2.6.1 Level-0 data products
     2.6.2 Level-1 data products
     2.6.3 Level-2 and higher data products
2.7 Example of Planetary Data Product Levels
     2.7.1 Level 1: Raw data
     2.7.2 Level 2: Raman datasets
     2.7.3 Level 3: Calibrated unidentified Raman spectra
     2.7.4 Level 5: Carbon/mineralogy results
     2.7.5 Level 6: Ancillary data
     2.7.6 Level 7: Correlative data
     2.7.7 Level 8: User description

3 Satellite Data and Image Quality Metrics
3.1 Needs for Quality Metrics
3.2 Full-Reference Metrics
     3.2.1 Conventional full-reference metrics
     3.2.2 Perceived-visual-quality-based full-reference metrics
3.3 Reduced-Reference Metrics
     3.3.1 Four RR metrics for spatial-resolution-enhanced images
     3.3.2 RR metric using wavelet-domain natural-image statistic model
3.4 No-Reference Metrics
     3.4.1 NR metric for compressed images using JPEG
     3.4.2 NR metric for pan-sharpened multispectral image

4 Satellite Data Compression
4.1 Lossless and Near-Lossless Data Compression
     4.1.1 Lossless compression
     4.1.2 Near-lossless compression
4.2 Vector Quantization Data Compression of Hyperspectral Imagery
     4.2.1 Review of fast VQ compression algorithms
     4.2.2 Near-lossless VQ compression techniques
4.3 Onboard Data Compression of Multispectral Images
     4.3.1 1D differential pulse code modulation
     4.3.2 Discrete-cosine-transform-based compression
     4.3.3 Wavelet-based compression
     4.3.4 Selective compression
4.4 Lossless Compression of Ultraspectral Sounder Data
     4.4.1 Comparison of wavelet-transform-based and predictor-based methods
     4.4.2 Lossless compression using precomputed vector quantization
     4.4.3 Lossless compression using the prediction-based lower triangle transform
4.5 CCSDS Data Compression Standards for Spacecraft Data
     4.5.1 Three space-data compression standards
     4.5.2 Lossless data compression standard
     4.5.3 Image data compression standard
     4.5.4 Lossless multispectral/hyperspectral compression standard

5 Satellite Data Formatting and Packetization
5.1 Formatting Satellite Data Using CCSDS Space-Data Link Protocol
5.2 Telemetry System Concept
     5.2.1 Packetization layer
     5.2.2 Transfer frame layer
     5.2.3 Channel coding layer
5.3 Space Packet Concept
5.4 Space Packet Structures
     5.4.1 Packet primary header
     5.4.2 Packet datafield
5.5 Telemetry Transfer Frame
     5.5.1 Transfer frame primary header
     5.5.2 Transfer frame secondary header
     5.5.3 Transfer frame datafield
     5.5.4 Operational control field
     5.5.5 Frame error control field

6 Channel Coding
6.1 Telemetry System Layers and Channel Coding
6.2 Channel Coding Improving Space Data Link Performance
     6.2.1 Channel coding performance measures
     6.2.2 Shannon limit on channel coding performance
6.3 Reed-Solomon Codes
     6.3.1 Definition
     6.3.2 RS encoder
     6.3.3 Interleaving of the RS symbols
     6.3.4 Decoding of RS codes
     6.3.5 Performance of RS codes
6.4 Convolutional Codes
     6.4.1 Encoder for CCSDS (7, 1/2) convolutional code
     6.4.2 Encoder for CCSDS punctured convolutional code
     6.4.3 Soft maximum-likelihood decoding of convolutional codes
     6.4.4 Performance of (7, 1/2) code and punctured convolutional codes
6.5 Concatenation of Reed-Solomon and Convolutional Codes
6.6 Turbo Codes
     6.6.1 Definition
     6.6.2 Turbo encoder and decoder
     6.6.3 Comparing turbo codes to traditional concatenation codes
6.7 Low-Density Parity-Check Codes
     6.7.1 Introduction
     6.7.2 CCSDS-recommended LDPC codes
     6.7.3 Performance of LDPC code

7 Calibration of Optical Sensors
7.1 Importance of Calibration
7.2 Absolute and Relative Radiometric Calibration
7.3 Satellite Optical Sensor Modeling
7.4 On-Ground Calibration prior to Launch
     7.4.1 Review
     7.4.2 Landsat instrument laboratory calibration
     7.4.3 AVIRIS laboratory calibration
7.5 Onboard Calibration Postlaunch
7.6 Vicarious Calibration
7.7 Conversion to At-Sensor Radiance and Top-of-Atmosphere Reflectance
     7.7.1 Conversion to at-sensor radiance
     7.7.2 Conversion to top-of-atmosphere reflectance
     7.7.3 Conversion to at-sensor brightness temperature

8 Keystone and Smile Measurement and Correction
8.1 Keystone and Smile in Imaging Spectrometers
     8.1.1 Spectral distortion: smile
     8.1.2 Spatial distortion: keystone
     8.1.3 How keystone and smile affect pixel shape and location
8.2 Method of Measuring Smile Using Atmospheric-Absorption Feature Matching
8.3 Smile Measurements of Five Hyperspectral Imagers
     8.3.1 Testing AVIRIS sensor smile
     8.3.2 Smile measurement of the SFSI sensor
     8.3.3 Smile measurement of the CASI sensor
     8.3.4 Smile measurement of the CHRIS sensor
     8.3.5 Smile measurement of Hyperion
8.4 Measuring Keystone Using Interband Correlation of Spatial Features
8.5 Measuring Keystone of Hyperspectral Imagers
     8.5.1 Test of keystone of AVIRIS sensor
     8.5.2 Measuring keystone of the Aurora sensor
     8.5.3 Measuring keystone of the CASI sensor
     8.5.4 Measuring keystone of the SFSI sensor
     8.5.5 Measuring keystone of the Hyperion sensor
     8.5.6 Summary of keystone measurement results
8.6 Effects of Keystone on Spectral Similarity Measures

9 Multisensor Image Fusion
9.1 Image Fusion Definition
9.2 Three Categories of Image Fusion Algorithms
9.3 Conventional Image Fusion Methods
     9.3.1 IHS fusion
     9.3.2 PCA fusion
     9.3.3 Arithmetic combination fusion
     9.3.4 Wavelet transform fusion
9.4 Comparison of Typical Image Fusion Techniques
     9.4.1 Brief description of nine fusion techniques
     9.4.2 Summary of evaluation results
9.5 Image Fusion Using Complex Ridgelet Transform
     9.5.1 Purpose
     9.5.2 Radon transform
     9.5.3 Ridgelet transform
     9.5.4 Operation of iterative back-projection
     9.5.5 Image fusion based on the complex ridgelet transform
     9.5.6 Image fusion experimental results
9.6 Fusion of Optical and Radar Images
     9.6.1 Fusion of multispectral and SAR images using intensity modulation
     9.6.2 SAR and optical image fusion based on wavelet transform
     9.6.3 SAR and optical image fusion based on local variance and mean
     9.6.4 Fusion of RADARSAT-1 and SPOT images

10 Enhancing the Spatial Resolution of a Satellite by Exploiting the Sensor's Keystone Distortion
10.1 Enhancing Satellite Sensor Performance Using a Signal Processing Approach
10.2 Exploiting the Keystone of a Satellite Sensor to Enhance Spatial Resolution
10.3 Using Keystone to Increase the Spatial Resolution of a Single-Band Image
     10.3.1 Fusion of subpixel-shifted images
     10.3.2 Method 1: Separate band images extracted based on KS-induced subpixel shift
     10.3.3 Method 2: Synthetic images derived based on a given amount of subpixel shift
     10.3.4 Method 3: Synthetic images derived based on closeness of pixel intensity
     10.3.5 Two schemes of organizing subpixel-shifted images and IBP implementation
10.4 Experimental Results of Single-Band High-Resolution Images
     10.4.1 Image quality metric: modified visual information fidelity
     10.4.2 Test hyperspectral datacubes
     10.4.3 Results of the Target datacube
     10.4.4 Results of the Key Lake datacube
10.5 Increase the Spatial Resolution of an Entire Datacube
10.6 Experimental Results of a Datacube after Spatial Resolution Enhancement
10.7 Conclusion and Discussion

11 Increasing the Signal-to-Noise Ratio of Satellite Sensors Using Digital Denoising
11.1 Increasing the SNR of Satellite Sensors by Reducing Noise
11.2 Hybrid Spectral�Spatial Noise Reduction
     11.2.1 Wavelet-shrinkage noise reduction
     11.2.2 Problem definition
     11.2.3 Proposed approach
     11.2.4 Experimental results of noise reduction
11.3 Noise Reduction Using Principal Component Analysis and Wavelet Shrinkage
     11.3.1 Combined PCA and wavelet-transform denoising method
     11.3.2 Test results of the combined PCA and wavelet denoising method
11.4 Combining Principal Component Analysis with Block-Matching 3D Filtering
     11.4.1 Combined PCA and BM3D denoising method
     11.4.2 Test results
11.5 Evaluation of the Hybrid Spectral-Spatial Noise Reduction Technique
     11.5.1 Remote sensing products used for the evaluation
     11.5.2 Evaluation criteria
     11.5.3 Evaluation results

12 Small-Target Detection of Hyperspectral Images after Noise Reduction
12.1 Target Detection of Hyperspectral Images
12.2 Spectral-Angle-Mapper-Based Method
     12.2.1 Test dataset
     12.2.2 Target superficies estimation using the SAM approach
     12.2.3 Target superficies estimation results
12.3 Receiver Operating Characteristic Method
12.4 Target Detection Using Spectral Unmixing
     12.4.1 Spectral unmixing and target masks
     12.4.2 Evaluation criteria
     12.4.3 Target detection and evaluation results
12.5 Target Detection by Summing Pixels' Endmember Fractions
     12.5.1 Subpixel target detection
     12.5.2 Target detection and evaluation results
     12.5.3 Discussion and conclusion

13 Dimensionality Reduction of Hyperspectral Imagery
13.1 Review of Three Dimensionality-Reduction Methods and Band Selection
     13.1.1 Principal component analysis dimensionality reduction
     13.1.2 Wavelet dimensionality reduction
     13.1.3 MNF dimensionality reduction
     13.1.4 Band selection
13.2 Evaluation of Three Dimensionality-Reduction Methods and a Band-Selection Method
     13.2.1 Using endmember extraction
     13.2.2 Using mineral detection
     13.2.3 Using mineral classification
     13.2.4 Using forest classification
     13.2.5 Summary
13.3 Reducing Dimensionality Using Locally Linear Embedding
     13.3.1 Nonlinear dimensionality reduction using modified LLE
     13.3.2 Evaluation using EM extraction and mineral detection
13.4 Reducing Dimensionality Using Combined LLE and Laplacian Eigenmaps
     13.4.1 Combined LEE and Laplacian eigenmaps dimensionality reduction
     13.4.2 Test results using EM extraction
13.5 Bivariate Wavelet Shrinking and PCA Method
     13.5.1 Reducing dimensionality and noise using BWS+PCA
     13.5.2 Evaluation of BWS+PCA
13.6 Reducing Dimension and Noise Simultaneously Using Wavelet Packets and PCA
     13.6.1 WP+NS+PCA method
     13.6.2 Evaluation of WP+NS+PCA method

14 Fast Endmember Extraction Based on the Geometry of the Datacube
14.1 Mixing Pixels and Linear Spectral Unmixing
14.2 Endmember-Extraction Methods
     14.2.1 Overview
     14.2.2 N-FINDR
     14.2.3 Simplex growing algorithm
     14.2.4 Pixel purity index
     14.2.5 Iterative error analysis
     14.2.6 Automated morphological EM extraction
     14.2.7 Automatic target generation process/vertex component analysis
     14.2.8 Fully constrained least-squares linear unmixing
14.3 Fast EM-Extraction Algorithm in a Reduced Search Space
     14.3.1 Fast N-FINDR
     14.3.2 Simulation results
     14.3.3 Discussion


Over the last two decades, I - a senior research scientist and technical authority with the Canadian Space Agency - have led and carried out research anddevelopment of advanced space technology in collaboration with my colleaguesat the agency and other government departments, my postdoctoral visitingfellows, internship students, and engineers in the Canadian space industry. We developed and patented a variety of novel signal processing methodologies and technologies for optical satellites. I was frequently invited by professors at universities (mostly in Canada) to give lectures to students; as a former professor myself, I�ve always enjoyed interacting with students and attempting to answer their questions. I was deeply touched by their eagerness and passion for acquiring knowledge and solving problems. In modern times, email is a powerful communication means: I often received emails from students around the world asking me to respond to their inquiries about my published works and to supply them with reference documents for their graduate work.

Although I have published over a hundred papers and currently hold nine U. S. patents, three European patents, and several pending patents in the subjects of optical satellite signal processing and enhancement, I have not previously organized these works into a book. This text ismy attempt to provide an end-to-end treatment of optical satellite signal processing and enhancement based on my 30 years of firsthand experience and research. It serves as an introduction for readers who are willing to learn the basics and the evolution of signal processing for optical satellites, and a guide for those working on the satellite image processing, data distribution, and the manipulation and deployment of satellite communications systems. The writing style provides clear and precise descriptions for advanced researchers and expert practitioners as well as for beginners. The structure of the chapters adopts a layout similar to journal papers, opening with a brief introduction on the subject matter, then reviewing previous approaches and their shortcomings, next presenting the recent techniques with improved performance, and finally reporting experimental results for assessing their effectiveness and providing conclusions.

Readers need not begin at the first page of the book and perform a sequential reading, but it is advisable to read Chapters 1 to 3 first; they cover the basics of spaceborne optical sensors, satellite data generation, and image quality metrics for assessing satellite images. Chapter 4 constitutes a separate part devoted to the topic of onboard satellite data compression. (For a more-comprehensive description of satellite data compression, readers are recommended to read Optical Satellite Data Compression and Implementation (SPIE Press, 2013), also by this author.) Chapters 5 - 8 constitute another part devoted to the subsequent processes of the data communication and calibration after the onboard compression has occurred, namely the transmission from the satellite to the ground, and then the calibration to remove the artifacts of the instrument. Chapters 9 - 14 constitute the third part, devoted to image enhancement and exploitation. Data is now available on the ground, and specialists are expected to derive qualitative application products. Processes for improving the quality of the available data and techniques to employ such data are presented. Instead of designing and building novel expensive payloads, cheaper signal processing algorithms are applied to reduce noise and increase the signal-to-noise ratio, spatial resolution, and other data characteristics.

There are many people I would like to thank for their contributions to the works included in this book. I would like to thank the Canadian Space Agency, where I have been working for the last 20 years; my colleagues Allan Hollinger, Martin Bergeron, Michael Maszkiewicz, Davinder Manak, and Ian Cunningham for their participation in data compression projects; the postdoctoral visiting fellows who I supervised, including Guangyi Chen, Reza Rashidi-Far, Hisham Othman, Pirouz Zarrinkhat, Charles Serele, and Riadh Ksantini for their contributions to denoising, enhancing spatial resolution, dimensionality reduction, spectral unmixing, target detection, and data compression; and over 40 internship students who have each left their mark in contribution. I would like to thank Robert Neville (retired), Karl Staenz (now with the University of Lethbridge), and Lixin Sun at the Canada Centre for Remote Sensing for allowing me to include their work on keystone and smile detection and their correction in this book, and for collaboration on the Canadian hyperspectral program; Jos�e L�vesque and Jean-Pierre Ardouin at the Defence Research and Development Canada for their collaboration on target detection and enhancement of spatial resolution.

I thank David Goodenough at the Pacific Forestry Centre; John Miller and Baoxin Hu at York University, for providing datasets and for actively collaborating on hyperspectral applications and Bormin Huang of the Cooperative Institute for Meteorological Satellite Studies at the University of Wisconsin-Madison for discussing satellite data compression. I also would like to thank Penshu Yeh at the NASA Goddard Space Flight Center, Aaron Kiely at the Jet Propulsion Laboratory, Carole Thiebaut and Gilles Moury at the French Space Agency (CNES), and Raffaele Vitulli at the European Space Agency for the collaboration within the CCSDS in developing international spacecraft data standards and for their contributions to the CCSDS work included in this book. I would also like to thank the three anonymous reviewers for their tireless work and strong endorsement of this book, their careful and meticulous chapter-by-chapter review on behalf of SPIE Press, and their detailed comments leading to the improvement and final results of the book in its current form. Many thanks as well to Tim Lamkins, Scott McNeill, and Dara Burrows at SPIE Press for turning my manuscript into this book.

Finally, this book would not have been possible without the help and support of my wife Nancy and daughter Cynthia, who provided great encouragement and assistance during the many hours of my spare time after work when I was preparing, typing, and editing this book. I owe great thanks to them for their patience and love.

Shen-En Qian
Senior Scientist, Canadian Space Agency
Montreal, Canada
September 2013

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research