| 摘要 |
This object recognition device comprises: a point cloud storage unit that stores information on point clouds which are sets of points; an input/output unit for inputting/outputting information related to said point cloud information; a shape recognition unit that estimates the shape attribute of said point cloud; and an object recognition unit that recognizes the presence of a target object in said point cloud information. Said point cloud information includes: information on a measurement point cloud which is an input of measurement values; and information on a model point cloud which corresponds to the target object to be recognized. Said shape recognition unit estimates, for each point in said measurement point cloud and said model point cloud, a shape attribute determined by relative coordinate information of points in the vicinity of each said point. Said object recognition unit performs a matching process for each said shape attribute of said measurement point cloud and said model point cloud that have the same shape attribute, and acquires information that said model point cloud is present in said measurement point cloud on the basis of the degree of matching between said shape attributes. |
