| Abstract | Not so long ago, a hill top view was the largest vista from which to observe nature's workings. Discoveries in optics, photography and flight have allowed us to see the Earth as never before. Advanced
methods in computing and signal processing technologies et cetera have enabled us to increase our ability to visualize and perceive the Earth's surface. Today, Earth observation satellites orbit our planet collecting data needed to produce images
which allow us to monitor, understand and plan the use of our world's resources. Throughout history, humans have tried to represent what they saw and understood through images. Everything from cave walls, to canvases, to computer screens have
been used to express perception of our surroundings. Maps have been one way to show the relationship between humans and their environment. Towns, roads, rivers, mountains, valleys, and where the land meets the sea, have been drawn in an organized
fashion for centuries. Mapmakers have always sought ways in which to represent both the location and the three dimensional shape of land. Mapmakers and other illustrators have traditionally used rendering techniques such as shading, overlapping
and perspective views to give an impression of three dimensionality. In the last 200 years, many advances in representing three dimensions have been made. Stereomodels, anaglyphs, chromo-stereoscopic images and holograms can provide three-dimensional
(3-D) information about our planet that flat, two-dimensional (2-D) images can not. Why is it important that the third dimension be conveyed? Humans are naturally able to see in three dimensions. The 'naturalness' of a 3-D representation of
reality enhances our ability to interpret 2-D imagery. Cartographers, engineers, geologists, hydrologists, and other scientists use 3-D viewing methods, such as stereo viewing of aerial photos and satellite images, in order to better understand the
Earth's surface. Representation of the third dimension supplies important information about relationships between land shape and structure, slopes, water ways, surface material and vegetative growth. Stereo viewing of two 2-D images has been
used since the mid 1800's. Stereoscopes are still widely used throughout the world. They are much less expensive and much more portable than computer hardware and software. Even though automatic methods to extract quantitative information from a
stereo pair have been developed, qualitative interpretation is best handled by people skilled in stereo viewing. For many users and applications, a stereoscope and a stereopair represent the most efficient way of getting a large amount of information
about an area of the Earth quickly and inexpensively. The following sections compose a training package designed to demonstrate the feasibility and potential of stereoscopy with respect to RADARSAT data. RADARSAT, Canada's first Earth observation
satellite, was launched in November 1995. It is a C-band SAR satellite with the ability to provide imagery of any part of the Earth's surface, in any climatic conditions, and by day or night. RADARSAT imagery is well suited to be used in stereo
because it can be collected from different look directions, beam modes, beam positions, and at resolutions fine enough to provide a good level of detail of the Earth's surface. This training package has been divided into six sections which are
intended to provide the background needed to fulfil the goal of this manual which is How to Use RADARSAT Data in Stereo. The manual is not intended to review and discuss in great detail the topics in each section. A bibliography has been provided for
those interested in pursuing any of the topics in greater detail. The first section provides background on 3-D visual ability. In this section, human visual perception with regards to both depth and colour is discussed. The second section is a
description of various methods used, historically and presently, to display 3-D information. The third and fourth sections discuss stereo and radar basics respectively. The fifth section is provides a short description of RADARSAT and it's ability to
acquire data specifically for stereo usage. Lastly, examples of RADARSAT and other remotely sensed data are used to illustrate concepts discussed in the previous sections and to showcase RADARSAT imagery in stereo. This section provides users of this
manual with hands on experience and should enable them to decide on how to generate the best stereo pair for a given application. |