Abstract
Cameras and inertial sensors are good candidates to be deployed together for autonomous vehicle motion estimation, since each can be used to resolve the ambiguities in the estimated motion that results from using the other modality alone. We present an algorithm that computes optimal vehicle motion estimates by considering all of the measurements from a camera, rate gyro, and accelerometer simultaneously. Such optimal estimates are useful in their own right, and as a gold standard for the comparison of online algorithms. By comparing the motions estimated using visual and inertial measurements, visual measurements only, and inertial measurements only against ground truth, we show that using image and inertial data together can produce highly accurate estimates even when the results produced by each modality alone are very poor Our test datasets include both conventional and omnidirectional image sequences, and an image sequence with a high percentage of missing data.
Highlights
Cameras and inertial sensors are each good candidates for autonomous vehicle navigation because they do not project any detectable energy into the environment, estimate six degree of freedom motion, are not subject to outages or jamming, and are not limited in range
We present an algorithm that instead considers all of the measurements from images, a rate gyro, and an accelerometer simultaneously to produce an optimal estimate of the vehicle motion and motion error covariances
Our method is a batch algorithm that uses all of the observations from an image sequence, rate gyro, and accelerometer to produce an optimal estimate of the sensor motion and the motion error covariances
Summary
Cameras and inertial sensors are each good candidates for autonomous vehicle navigation because they do not project any detectable energy into the environment, estimate six degree of freedom motion, are not subject to outages or jamming, and are not limited in range. Cameras and inertial sensors are good candidates to be deployed together, since in addition to the obvious advantage of redundant measurements, each can be used to resolve the ambiguities in the estimated motion that results from using the other modality alone. We present an algorithm that instead considers all of the measurements from images, a rate gyro, and an accelerometer simultaneously to produce an optimal estimate of the vehicle motion and motion error covariances. In many applications, this optimal estimate is of interest in its own right. Optimal estimates are useful as a gold standard for the comparison of online algorithms which, given sufficient computing power, produce real-time but suboptimal estimates of the vehicle’s motion
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have