jdye@ADS.COM (John W. Dye Jr.) (10/09/90)
I have a lot of "range" images taken from a laser radar imager which I need to be able to visualize in as close to 3 dimensions as possible. These are images of outdoor scenes in fairly flat landscape with objects in the landscape providing most of the interesting features. The images are valued 0-256 corresponding to one complete phase cycle of the returned signal. Because they only use phase information to create the image, the images have an inherent range ambiguity which causes a banding effect as the terrain falls away from the sensor. This banding would need to be removed in any type of surface rendering. Typically these images are displayed as a greyscale 2d image and the grey values represent the distance from the sensor. This gives a rough image of "sillouettes" of the objects in the scene. We would like to be able to do a 2 1/2 dimensional "surface" display with these objects. Is anyone in netland experienced with visualizing these images? Is there any public domain software (part of an imaging package, maybe?) for handling this task. In particular, we need software to: 1) Eliminate the "banding" due to phase ambiguity. 2) Map the distance (from sensor) values into a different value space (say X,Y,Z). 3) Display the converted values as a surface grid (ala mathematica or terrain surface displays). 4) Possibly perform a texture map of the original pixel values onto the surface grid. Any help or pointers would be appreciated. Thanks, JD jdye@ads.com