Rover-Based Visual Target Tracking Validation and Mission Infusion

the art in rover operations on Mars. To place an instrument on 10 to 20 m target designated by scientists on Earth, the MER baseline operations require 3 sols (Martian days). In future Martian surface operations, it is desirable to achieve the entire 10-m target approach and instrument placement in a single sol, increasing the science return significantly. One of the key technologies to achieve this capability is visual target tracking during the rover approach towards the designated target. This paper presents validation experiments of different visual tracking algorithms using the rover’s navigation camera (navcam). Normalized cross-correlation (NCC) matching with template image scaling (magnification) performed best in terms of tracking reliability and accuracy. Although affine tracking tended to be more accurate when it worked, it was less reliable. Various navcam tracking experiments with the scaled NCC matching were performed and analyzed for straightforward, sideways, and turn-in-place rover motions to determine the appropriate motion step sizes. The paper also presents on-going efforts of infusing the visual target tracking software written and tested in the CLARAty software environment into the MER flight system. The ported visual target tracking software will be tested on Mars as one of several technology demonstrations planned for MER extended flight experiments. I