Image processing for precise geometry determination

TL;DR

This paper presents a comprehensive image processing pipeline that corrects spatial metadata errors and geometric distortions in planetary images to accurately determine the geometry of observed features.

Contribution

The authors develop a novel pipeline that combines correction of metadata errors, distortion adjustments, and efficient pixel projection for precise geometry determination in planetary remote sensing images.

Findings

Effective correction of spacecraft position and attitude errors.

Accurate projection of image pixels onto elliptic targets.

Improved geometric measurements of planetary surface features.

Abstract

Reliable spatial information can be difficult to obtain in planetary remote sensing applications because of errors present in the metadata of images taken with space probes. We have designed a pipeline to address this problem on disk-resolved images of Jupiter's moon Europa taken with New Horizons' LOng Range Reconnaissance Imager, Galileo's Solid State Imager and Voyager's Imaging Science Subsystem. We correct for errors in the spacecraft position, pointing and the target's attitude by comparing them to the same reference. We also address ways to correct for distortion prior to any metadata consideration. Finally, we propose a vectorized method to efficiently project images pixels onto an elliptic target and compute the coordinates and geometry of observation at each intercept point.