Depth Estimation Using Particle filters for Image-based Visual Servoing

Abstract

In this paper, we present novel method for depth estimation in image-based visual servoing.This method employ the particle filter algorithm to estimate the depth of the image features on-line. AGaussian probabilistic model is employed to model the depth. A set of depth particles is drawn in thecurrent camera frame. The image measurements are used to recover the 3D samples. These samples arepropagated to the next frame and projected into the image space. The maximum likelihood of the 3Dsamples is the most probable to be the real-world 3D point. The mean and the variance of the depthdistribution is obtained from the maximum likelihood. The variance values converge to very small valuewithin a few iterations. This gives high level of stability to the image-based visual servoing system.The simulation experiments show that the mean goes to the real value of the depth in a few iterations.The depth is considered as the mean value of the estimate. Alternatively, result in a depth estimate also,generating random sample of the estimated distribution then substitute it in the control law.