Reliable Planar Object Pose Estimation in Light Fields From Best Subaperture Camera Pairs

A light-field camera can obtain richer information about a scene than a usual camera. This property offers a lot of potential for robot vision. In this letter, we present a method for pose estimation of a planar object with a light-field camera. The light-field camera can be regarded as a set of subaperture cameras. Although any combination of them can theoretically be used for the pose estimation, the accuracy depends on the combination. We show that the estimated pose error can be reduced by selecting the best pair of subaperture cameras. We have evaluated the accuracy of our approach with real experiments using a light-field camera in front of planar targets held by an industrial manipulator for ground truth comparison.