Space Image Processing

Julio Sanchez

• Publisher: Routledge
• Year: 2018
• Print ISBN: 9780849331138, 9780849331138
• eBook ISBN: 9781351415422
• Edition: 1
• Topic: Technology & Engineering
• Subtopic: Electrical Engineering & Telecommunications
• Index: Technology & Engineering

Space Image Processing covers the design and coding of PC software for processing and manipulating imagery obtained by satellites and other spacecraft. Although the contents relate to several scientific and technological fields, it serves as a programming book, providing readers with essential technical information for developing PC applications. The material focuses on images of the planet and other celestial bodies obtained by orbiting and non-orbiting spacecraft. This book is not about raster graphics in general, but about raster graphics processing as it applies to space imagery.

Three parts divide the text: 1. Science - background at an introductory level - scientific principles underlying space imagery and its processing - topics related to space and remote sensing. 2. Technology - topics related to space imagery - geodesy, cartography, image data formats, image processing. 3. Programming - code examples for DOS and Windows programming on the PC - consideration of low-level and C++ code - routines with a tutorial and demonstrative purpose.

PART I

The Science

Chapter 1

Remote Sensing

1.0 Introduction

Remote sensing is defined as the art and science of obtaining information about an object by a device that is not in direct contact with it. In a sense most information is obtained by remote sensing since even our eyes are not in direct contact with the objects we see. However, in a technological context, remote sensing usually refers to data gathered by sensors and instruments that measure emitted or reflected electromagnetic radiation which is formatted digitally so it can be viewed pictorially or analyzed by computers.

But, even when the notion of remote sensing is restricted to instruments that detect and measure electromagnetic radiation, the resulting concept is still quite broad. In this definition, no restriction is placed on the sensor instrument itself, on the object being sensed, or on the location of the sensor. Therefore, this notion of remote sensing includes digital cameras used in conventional photography, radiological and other sensor devices used in medicine, and many other specialized fields. A more conventional definition restricts remote sensing to Earth remote sensing. This refinement usually focuses on Planet Earth as the object being sensed and assumes that the sensing instrument is either on board an aircraft or a satellite. However, there is no scientific or technological reason for restricting remote sensing to our planet. A satellite orbiting a planet or other celestial body can generate images that are very similar to those produced by a similar instrument orbiting Earth. Furthermore, much of the knowledge acquired in Earth remote sensing is also applicable to “planetary” remote sensing.

1.1 Earth Remote Sensing

In the more conventional sense adopted in the preceding section, remote sensing uses instruments mounted on airplanes and satellites to acquire digital images of the Earth’s surface. The most familiar versions of remotely sensed images are those furnished by the weather segments of televised news programs. However, during the past thirty years, remote sensing has been used in many fields of investigation and research, such as mineral exploration, monitoring ocean temperatures and currents, detecting changes in the Earth’s biosphere, as support for geographic information systems, observing types of vegetation and their state of health, in the scrutiny of rocks and minerals on the Earth’s surface, measuring water depths and other hydrological features of the geosphere, and in detecting land cover and planning land use.

Perhaps the most unique attribute of this technology relates to the possibility of showing large land areas of the Earth’s surface and in detecting features that are not visible to the human eye. The fact that satellites can be made to pass over the same area at regular intervals allows recording and monitoring changes in the oceans and land masses. The sequence of images that can be obtained over time provide an additional dimension to this resource.

1.1.1 Related Fields

The images obtained by satellites and spacecraft are varied in nature. In the first place, the imagining instrument can be aimed toward earth, or toward any other part of the universe. In the context of remote sensing we concentrate on images obtained by earth-orbiting and earth-observing satellites and let astronomers and planetary scientists deal with images of other celestial bodies. But even with this restriction, the target object (Planet Earth) is of interest to a multitude of scientific fields, for example,

1. A satellite image contains geographical information that is useful for creating or updating charts or maps. In this case, the image processing operations refer to the geophysical sciences, such as geodesy and cartography. Modern day Geographic Information Systems (GIS) rely heavily on remotely-sensed data for periodic corrections and updates.
2. Plants, terrestrial and aquatic, are a substantial part of the earth’s surface, therefore, satellite images of the Earth are of interest to plant science. The detailed investigation of plant life is possible because the living leaf shows a unique spectral behavior. For this reason, remotely sensed images are used in classifying and mapping vegetation, in identifying ecosystems, in measuring forested areas, in identifying crops and rangelands and detecting their health, in detecting and measuring seasonal changes of the canopy, and in many other applications of forestry and agriculture.
3. Another element of a remotely-sensed image is the earth’s surface itself. Consequently, the earth sciences (geology, geomorphology, and soil science) also find interest in this material. Applications in these fields are complicated by the fact that many geological features are partly hidden by vegetation or located beneath the earth’s surface. Nevertheless, remotely sensed images give geo-scientists a broad-scale perspective that is not available from ground observations, as well as a measurement of the reflected and emitted wavelengths of large-scale geological objects.
4. Approximately 74 percent of the earth’s surface is water, of which oceans account for about 95 percent. Freshwater, a proportionately small part of the earth’s total water, is one of mankind’s most important resources. Precipitation, lakes, rivers, and, in temperate zones, the seasonal melting of the snowpack are our major sources of moisture. Hydrologists and meteorologists use satellite images to monitor oceans, ice masses, lakes, rivers, and snowpack. The spectral characteristics of water bodies allow measuring their composition and depth, as well as determining their location and extent. Chromatography (measurement of color) allows analyzing the turbidity and composition. Multispectral bathymetry allows estimating depths.
5. Humans use land in their economic activities: in creating cities and suburbs, in mining and industry, and in agriculture. We cover the land with cement, with asphalt, with industrial and human debris, and with crops. Land use and land cover indicate patterns of interaction between a society and its environment. How land use changes over time is of interest to governments, to commercial firms and other organizations, and to us as individuals. Remotely sensed images offer a powerful means for detecting and measuring the use of this ever-decreasing resource.
6. We have not yet precisely determined the effect of human activities on the tender, life-supporting shell of our planet. The images obtained by satellites depict the earth’s biosphere, therefore, they can be used in studying the earth’s patterns of climate and biology. These associated scientific activities, sometimes called global remote sensing, may very well turn out to be the most transcendental application of the technology.

This list, although incomplete, gives an idea of the scientific and technological extent of Earth remote sensing. Any one of the mentioned fields or subfields of specialization can easily constitute the life-time interest of a scientist, investigator, or specialist. By necessity, our treatment of any one of these specialty topics will be partial and superficial.

1.1.2 Technical Fundamentals

Remote sensing instruments measure the intensity of emitted or reflected radiation at different wavelengths. The fundamental application of this technology assumes that particular objects and geological features reflect and emit radiation in different wavelengths and strengths. In this respect, plants appear green because they reflect more green light than the other visible colors. Furthermore, some healthy plants strongly reflect radiation in the infrared region of the spectrum. Although we cannot see infrared radiation, it is possible to capture and record it with a sensor. The resulting infrared record can then be displayed as a visible image.

Color plate 1 shows two interpretations of a satellite image. The first image, labeled a, is an approximation of what we would see if we were looking at this area, with our naked eyes, from the satellite. The second image, labeled 6, is a color manipulation of three spectral bands, one in the visible range and two in the infrared region, in order to show features that are not directly visible. A geologist may be able to draw information about rock formations and mineral deposits from this interpretation of the data.

Remote sensing is a relatively new field: many of its fundamental facts are not yet known. Every year new investigative methods and interpretation of exiting data are discovered. However, there are certain scientific certainties and research principles that appear to be fundamental to the technology, for example,

1. The data obtained by sensors is meaningful because there are detectable differences in the spectral response of different objects, regions, or landscape features. This fact is the basis of multi-spectral remote sensing, which separately detects different wavelength areas (usually called spectral bands) and uses this data to capture information about an object, region, or feature. The term spectral signatured often used in this context to describe the particular response of a specific object. However, the interpretation of spectral signatures is complicated by natural variations in the object itself as well as by climatic and atmospheric conditions. In this sense, there is no such thing as the spectral signature of a wheat field: the actual values change with the state of health of the crop, its age, the season, the atmospheric conditions at the time of the sensing, and other factors.
2. Every sensor is limited by the size of the smallest area that can be recorded and stored. We speak of a sensor’s resolution as...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• Introduction
• PART I - The Science
• PART II - The Technology
• PART III - The Programming
• Appendix A — ΤΜ-Lab Program Manual
• Bibliography
• Index