| 摘要 |
PROBLEM TO BE SOLVED: To provide the technology for simply performing individual identification of an object. SOLUTION: A feature extraction part 23 detects each feature of a noticed object included in an image, a local area setting part 24 sets a local area including a group of features to obtain the object shape and a local area including a group of features to obtain positional relation among the detected features, a feature vector generation part 25 calculates feature vector data in the set local area and an object recognition part 26 specifies which of mutually different objects the noticed object is based on the feature vector. COPYRIGHT: (C)2006,JPO&NCIPI
|
