Detection of Planar Motion Objects Using Radon Transform and One-Dimensional Phase-Only Matched Filtering

摘要

We describe an algorithm to detect the position and orientation of multiple objects in planar motion using the Radon transform and 1D phase-only matched filtering (POMF). The proposed vision algorithm performs pattern matching between a template and input image to detect the position and orientation of the objects. First, the Radon transform is applied to the two images to separate the rotation and translation of the objects. Next, 1D POMF along the distance of the Radon transform is applied to the transformed images so as to detect the separated position and orientation. Since it uses 1D POMF to detect both position and orientation, the proposed algorithm can detect the position and orientation of multiple objects.