Two Dimensional Pose Recognition of Objects Using Geometrical and Gray-levels Features between Edge Pixels Pairs
Yosihumi Matuura, Fumihiko Saitoh · IEEJ Transactions on Electronics Information and Systems · 2005
This paper proposes a method to measure the two-dimensional pose of an object by using vote processing based on geometrical features between edge pixels pairs. The proposed method can measure the set of location, inclination and scale of an object even if the object has a free pose in an image. In the template registration process, the geometrical features and the gray-levels information about all edge pixels pair are registered into the template table. In the matching process, not only the geometrical features but also the arrays of gray-levels between the edge pixels pair are used for the matching. The reliable pose measurement may be expected even if the environment of illumination is changed because the method uses the normalized gray-levels information of the internal region of the object for matching. The experimental results show that the poses of objects with free locations, inclinations and scales were measured successfully by the proposed method.