Computer Vision on Mars

Abstract

Increasing the level of spacecraft autonomy is essential for broadening the reach of solar system exploration. Computer vision has and will continue to play an important role in increasing autonomy of both spacecraft and Earthbased robotic vehicles. This article addresses progress on computer vision for planetary rovers and landers and has four main parts. First, we review major milestones in the development of computer vision for robotic vehicles over the last four decades. Since research on applications for Earth and space has often been closely intertwined, the review includes elements of both. Second, we summarize the design and performance of computer vision algorithms used on Mars in the NASA/JPL Mars Exploration Rover (MER) mission, which was a major step forward in the use of computer vision in space. These algorithms did stereo vision and visual odometry for rover navigation and feature tracking for horizontal velocity estimation for the landers. Third, we summarize ongoing research to improve vision systems for planetary rovers, which includes various aspects of noise reduction, FPGA implementation, and vision-based slip perception. Finally, we briefly survey other opportunities for computer vision to impact rovers, landers, and orbiters in future solar system exploration missions.

Real-time one-dimensional motion estimation and its application in computer vision

“…Visual odometry is the technique to estimate the pose (position and orientation) of the camera from sequential input images, and it has been applied in unmanned aerial vehicle (UAV) [15], Mars Rover [48] and unmanned ground vehicle (UGV) [51,57]. The key technology for VO is motion estimation, and two common technologies are general optical flow [16] and sparse invariant feature points [8].…”

Visual odometry aided by a sun sensor and inclinometer

“…Thus, in the interests of covering maximal ground, the use of VO has been limited to short drives (less than 15 m) involving specific terrain and tasks, despite its high positional accuracy. While the upcoming Mars Science Lab (MSL) rover employs a faster 200 MHz processor, its effective driving speed will still be limited by the VO update cycle time (Johnson et al, 2008;Matthies et al, 2007).…”

Mining remote-image repositories with application to Mars Rover stereoscopic image datasets

“…Of particular interest for geological research is the 3D stereoscopic data produced by the Pancam imaging system. This data is generated via a technique known as stereoscopic 3D reconstruction and is stored in a XYZ Reduced Data Record (RDR) which contains estimates for 3D (x,y,z) surface positions of objects viewed within a stereoscopic pair of Pancam EDR images [16]. The underlying concept of 3D reconstruction from images taken from a pair of cameras is that by knowing or estimating the image formation properties of each camera, their relative pose, and the pixel pair in each digital image that corresponds to a specific 3D surface location, one may invert the image formation process and find the 3D locations responsible for reflecting the light sensed by the camera.…”