| 摘要 |
The present invention provides a method of compiling an image database for a three-dimensional object recognition comprising the steps of: when a plurality of images each showing a three-dimensional object from different viewpoint are inputted, extracting local features from each of a plurality of images, and expressing the local features using respective feature vectors; forming sets of the feature vectors, each set representing the local feature of a same part of the object from a series of the viewpoints neighboring one another, and generating subspaces using a subspace method, each subspace representing a characteristic of each set; and storing each subspace to the image database in association with an identifier of the object, wherein the image database is accessed by a computer to perform a recognition process of the three-dimensional object, and the recognition process is realized by the steps of: when one image of an object from a certain viewpoint is given as a query, extracting a plurality of feature vectors as query feature vectors representing local features of the query; determining the subspace most similar to each query feature vector; and executing a counting process to the identifiers, each identifier being associated with each subspace, to retrieve a three-dimensional object most similar to the query image. |
