Three-dimensional target multi-viewpoint view modeling method based on support vector data description

Abstract

The invention provides a three-dimensional target multi-viewpoint view modeling method based on support vector data description. The method comprises three steps of 1, obtaining a full gesture image of a test target; 2, extracting a target characteristic vector set {xi|i=1,2,ellipsis N} from the obtained full gesture image set; and 3, representing the distribution of the characteristic vectors extracted from the image in a characteristic space by a support vector data description (SVDD) method, and using the obtained few characteristic vectors with fully representativeness as the multi-viewpoint view modeling result of the target. Aiming at the problem that the single target view description cannot identify the target caused by difference of the target image under different viewpoints, the invention discloses the three-dimensional target multi-viewpoint view modeling method with less redundant information and capability of describing the full gesture characteristic vector of the target. The method provided by the invention has good practical value and wide application prospect in the field of pattern recognition.

Description

technical field [0001] The invention relates to a three-dimensional target multi-view point view modeling method based on support vector data description, belongs to the field of pattern recognition, and specifically relates to the fields of target recognition, target modeling, data reduction and the like. It is used for multi-viewpoint modeling of 3D targets, and is suitable for the problem that the single target view description cannot recognize the target caused by the difference of target images under different viewpoints. Background technique [0002] 3D object recognition is an important research direction in the field of computer vision. At present, it is often very difficult to obtain the three-dimensional information of the target in practical applications, and the recognition of the three-dimensional target is still mainly completed by recognizing the image formed by the two-dimensional projection of the target. The process of two-dimensional imaging (projection) ...

AI Technical Summary

Problems solved by technology

[0007] The object of the present invention is to provide a 3D object multi-viewpoint view modeling method based on support vector data description, which aims at the disadvantage that the single object view description cannot identify the target caused by the difference of the target image under different viewpoints, and constructs a description quantity A 3D object multi-viewpoint view modeling method with small size, less redundant information, and better description of the object's full pose feature vector set

Images