Method and apparatus for generating shape descriptor of a model

Patent metadata

Field	Value
Publication number	US-9953458-B2
Application number	US-201214402186-A
Country	US
Kind code	B2
Filing date	May 22, 2012
Priority date	May 22, 2012
Publication date	Apr 24, 2018
Grant date	Apr 24, 2018

Abstract

The invention provides a method for generating an n dimensional vector as a shape descriptor of a model, and corresponding apparatus and shape descriptor. The method comprises: determining a type element of the vector to describe the basic shape of the model; and calculating n−1 metric elements of the vector, each of which represents the percentage of all of a feature of the model falling into one of n−1 layers divided as a function of the type element.

First claim

The invention claimed is:

1. A computer implemented method for generating an n dimensional vector as a shape descriptor of a non-watertight three-dimensional (3D) model, comprising:
determining, as a function of three eigenvalues of a covariance matrix of the non-watertight 3D model, a type element of the vector to describe a basic shape of the model, wherein the type element is a first dimension in the vector;
determining a feature of the non-watertight 3D model and a total number of features in the non-watertight 3D model;
generating n−1 metric elements of the vector, each of the n−1 metric elements represents all of the feature of the non-watertight 3D model falling into one of n−1 layers divided as a function of the type element per the total number of features in the non-watertight 3D model; and
performing a search for the non-watertight 3D model using the vector which can effectively process non-watertight 3D models.

2. The method according to claim 1 , wherein the feature comprises a vertex, a triangle, or an edge of the non-watertight 3D model.

3. The method according to claim 1 , wherein the basic shape of the non-watertight 3D model comprises cube, sphere and cylinder.

4. The method according to claim 1 , wherein the type element is an integer.

5. The method according to claim 1 , wherein a bounding box of the non-watertight 3D model is divided by n−1 embedded shape determined according to the type element.

6. The method according to claim 5 , wherein one of the metric elements is calculated by a normalization of a percentage of all of a feature of the non-watertight 3D model falling into one of n−1 layers.

7. The method according to claim 1 , wherein the metric elements are all floats.

8. Apparatus for generating an n dimensional vector as a shape descriptor of a non-watertight three-dimensional (3D) model, comprising:
a processor for determining, as a function of three eigenvalues of a covariance matrix of the non-watertight 3D model, a type element of the vector to describe the basic shape of the model, wherein the type element is a first dimension in the vector, said processor further configured to;
determine a feature of the non-watertight 3D model and a total number of features in the non-watertight 3D model;
the processor is further used to generate n−1 metric elements of the vector, each of the n−1 metric elements represents all of the feature of the non-watertight 3D model falling into one of n−1 layers divided as a function of the type element per the total number of features in the non-watertight 3D model; and
the processor is further used to perform a search for the non-watertight 3D model using the vector which can effectively process non-watertight 3D models.

9. The apparatus according to claim 8 , wherein the processor is used further for dividing a bounding box of the non-watertight 3D model with n−1 embedded shapes, the n−1 embedded shapes being determined according to the type element.

10. The apparatus according to claim 8 , wherein the processor is used further for determining one of the metric elements with a normalization of a percentage of all of a feature of the non-watertight 3D model falling into one of the n−1 layers.

Assignees

• Cai Kangying
• Meng Wei Liang
• Luo Tao
• Thompson Licensing Sa

Inventors

• Cai Kangying
• Meng Wei Liang
• Luo Tao

Classifications

• G06V10/46

  Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features (colour feature extraction G06V10/56) · CPC title

• G06T17/00 Primary

  Three-dimensional [3D] modelling for computer graphics · CPC title

• G06K9/48

  Physics · mapped topic

• G06K9/00201

  Physics · mapped topic

• G06T17/30

  Polynomial surface description · CPC title