My strategy was to use all seven bands of the SIR to create an interesting image of Sunbury Pennsylvania. The trick to doing this is to map the bands into numerous perceptual channels. I decided to use L_hv, C_hv, and X_vv for the red, green and blue channels, respectively. My selection was simply on the basis of attempting to map the wavelengths from longest to shortest into the optical band; red longest, blue shortest.
With the rest of the channels I wanted to attempt to generate an elevation map. This map probably has little correlation to the actual elevation of Sunbury, but could serve as a cue to the relative strength of the radar returns across those bands. Areas where there are strong returns are elevated (the city of Sunbury) and areas where there is little return are at low spots (the river for example).
Finally, although probably outside the scope of the contest, I decided to create a Quicktime movie and perform a fly through (took about a day to generate on a PowerPC). Our perceptual systems are very good at pulling information out of objects in motion.
Opened Adobe Photoshop v3.0 and loaded the L band hv into the red channel, the C band hv into the green channel and the x band vv into the blue channel.
L-band is 21 cm Red Channel hv
C-band is 5.6 cm Green Channel hv
X-band is 3 cm Blue Channel vv
Next I adjusted Brightness/Contrast Brightness -14, Contrast +28 Saved as Sunbury Texture in pict format
Opened c hh, l hh, c vv, l vv in multi channel mode assigning each a separate channel.
Created new channel Results c hh X l hh, used apply image to normal c hh then multiply l hh with 100% opacity set on both
Created new channel Results c hh X l vv, used apply image to normal c hh then multiply l vv with 100% opacity set on both
Created new channel Results, used apply image to normal Result c hh X l hh channel, used apply image overlay 50% opacity to Result c vv X l vv channel.
Applied Brightness +16 and Contrast +41 to the Results Channel saved pict file as SunburryHeight.
Since I intended to sample this image at every five pixels to derive elevation, I ran a Gaussian Blur to avoid sampling errors.
Gaussian Blur with a radius of 5 pixels converted results channel to grayscale and saved as SunburyHeight_blur
Opened Infini D 2.6 from Specular, Created a new Terrain object and scaled x and y by 10, centered at the origin. Under Render:Get Image... loaded Sunburytexture.pict, applied that as a texture map to the terrain.
Under Edit Terrain, changed the grid resolution to 100, selected image, edit image and loaded the sunburyheightblur.pict as the height map for the terrain.
Created a new sphere object, selected a uniform scale of 100, turn backfaces on, cast shadows off, then applied misty clouds as the texture map to the sphere.
Rendered Image at 640X480 no shadows, no dithering, no reflections, no transparency
Moved around to another view, scaled the terrain height (z scale in object info) to two - doubling the relative height. Rendered another 640X480 image.
Infini D is a powerful 3D modelling and visualization program from Specular Inc. One of the features that I like to use is the terrain object. The terrain object is a wire mesh that can be texture mapped and elevation mapped with imported pictures.I created an invisible cube and made it the parent object for the terrain. That allows separate rotation control of the cube and scaling control of the terrain and provides more flexibility for devising animation scenarios. Next, using the sequence window I created events that rotated the cube (and the attached terrain) at 90 degree intervals.
Navigating with the camera I moved to several positions along a flight path and saved them as snapshots. During this process the animation can be previewed in wireframe and test animations can be quickly generated by setting the resolution to 160 X 120, fast rendering, and five images per second. Once I was satisfied with the scenario I generated a Quicktime animation at 320 X 240, ray traced, 15 images per second.
Since the contest is restricted to jpeg or gif images, I read the pict files back into Adobe Photoshop and saved them as jpegs.Mark Lucas
Imaging Radar Home Page