Outdoor Downward-Facing Optical Flow Odometry with Commodity Sensors

Abstract

Positioning is a key task in most field robotics applications but can be very challenging in GPS-denied or high-slip environments. A common tactic in such cases is to position visually, and we present a visual odometry implementation with the unusual reliance on optical mouse sensors to report vehicle velocity. Using multiple kilometers of data from a lunar rover prototype, we demonstrate that, in conjunction with a moderate-grade inertial measurement unit, such a sensor can provide an integrated pose stream that is at times more accurate than that achievable by wheel odometry and visibly more desirable for perception purposes than that provided by a high-end GPS-INS system. A discussion of the sensor's limitations and several drift mitigating strategies attempted are presented.

An experimental study on using visual odometry for short-run self localization of field robot

“…By applying (1)(2)(3)(4) to images shown in Fig.1, phase correlation of two images can be obtained. At the location of maximum translation a impulse is created as shown in Fig.2.…”

“…Though similar techniques are used t robots [2] most of them are limited to in and data processing is done offline. Pra the proposed visual odometry system mai the ability of the algorithm and assoc accurately register two consequent frames…”

Improving Robot Mobility by Combining Downward-Looking and Frontal Cameras

“…The motivation of using one downward-looking camera in this research comes from [14,15]. In [14], a combination of four optical mouse sensors -and an INS-GPS module-is employed to perform visual odometry for a lunar rover. The vision module is mounted under the vehicle facing downwards, and it reports lateral motion velocities.…”

Optical flow odometry with robustness to self-shadowing

“…Feature selection, even though performed every iteration, does not contribute to latency because it is performed before rather than after the capture of I curr . (see Algorithm 1, lines [2][3][4][5] The top vehicle speed that can be tracked for a downwardlooking camera (mounted normal to the ground) is given by:…”