Method and device for selecting characteristics of three-dimensional objects
A feature selection method and three-dimensional object technology, applied in the multimedia field, can solve problems such as heavy workload, and achieve the effect of avoiding feature redundancy and reducing workload.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0024] see figure 1 , the present embodiment provides a feature selection method of a three-dimensional object, the specific process of the method is as follows:
[0025] 101: Respectively obtain similar class sets of three-dimensional objects under each feature;
[0026] 102: Select the smallest set of similar classes;
[0027] 103: Combine the features corresponding to the minimum similarity class set with other features to obtain a new combination feature, and obtain the similarity class set of the 3D object under the new combination feature until an empty similarity set is obtained. class set, using the new combined feature corresponding to the similar class set which is an empty set as the approximate optimal feature of the three-dimensional object.
[0028] The method provided in this embodiment, by making full use of the different features of the 3D object, obtains the similar class set of the 3D object under each feature, so as to obtain the approximate optimal featu...
Embodiment 2
[0030] This embodiment provides a method for feature selection of 3D objects. Before the feature selection of 3D objects, all 3D objects in the 3D object database are manually divided into several categories, and each category is marked; secondly, the Each 3D object is placed in the camera array for shooting, and a 2D view set corresponding to each 3D object is obtained, and these 2D view sets form a multi-view based 3D object database. The premise of implementing the method provided in this embodiment is: after feature extraction has been performed on the view set, if all the features of the 3D object are used, each 3D object should be able to be distinguished from other 3D objects of different categories from itself, if If the effect of distinction cannot be achieved, it is necessary to increase the characteristics of the 3D object or change the category label of the 3D object until this premise is met.
[0031] For the sake of illustration, take the feature selection of the...
Embodiment 3
[0059] see image 3 , the present embodiment provides a device for feature selection of a three-dimensional object, the device comprising:
[0060] An acquisition module 301, configured to acquire similar sets of three-dimensional objects under each feature;
[0061] Selection module 302, for selecting the minimum similar class set;
[0062] The processing module 303 is used to combine the features corresponding to the smallest similar class set with other features to obtain a new combined feature, and obtain the similar class set of the three-dimensional object under the new combined feature until an empty similar class set is obtained , will use the new combined feature corresponding to the similar class set of the empty set as the approximate optimal feature of the 3D object.
[0063] Further, see Figure 4 , the above acquisition module 301, including:
[0064] An extraction unit 301a, configured to extract each feature of the three-dimensional object;
[0065] Calcul...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 