The Effect of Incidence Angle on Stereo DTM Quality: Simulations in Support of Europa Exploration

Abstract:

Many quality factors for digital topographic models (DTMs) from stereo imaging can be predicted geometrically. For example, pixel scale is related to instantaneous field of view and range, and vertical precision is a known function of the pixel scale and convergence angle if the image quality is high so that automated image matching reaches its optimal precision (~0.2 pixel). The influence of incidence angle is harder to predict. Reduced quality is expected both at low incidence (where topographic shading disappears) and high incidence (where signal/noise ratio is low and shadows occur). This problem is of general interest, but critical for NASA’s Europa mission. The REASON instrument will obtain a radar sounding profile on each Europa flyby. Stereo images collected simultaneously by the EIS camera will be used to make DTMs needed to distinguish off-nadir surface echos (clutter) from subsurface features. The question is, how much of this DTM strip will be useful, given that incidence angle will vary substantially?

We are using simulations to answer this question. We mosaicked 220 m/pixel Galileo images of the Castalia Macula region of Europa and made a DTM by photoclinometry, using a low-incidence image to correct for albedo variations. With the simulation software OASIS we generated synthetic stereopairs that are realistic in terms of image resolution, noise, photometry (including albedo variations based on the low incidence image), and cast shadows. We then use the commercial stereo software SOCET SET (® BAE Systems), which we have used for a wide variety of planetary mapping projects, to produce DTMs. These are compared to the input DTM as “truth” to quantify the dependence of DTM resolution and vertical precision on illumination, and to document the ways that DTMs degrade at high and low incidence angles.

This methodology is immediately applicable to other planetary targets, and in particular can be used to address how much difference in illumination can be tolerated in stereopairs that are not (as for EIS) acquired simultaneously.