Method and system for cleaning motion capture data

• US 7,324,110 B2
• Filed: 12/02/2005
• Issued: 01/29/2008
• Est. Priority Date: 12/09/2004
• Status: Active Grant

• Current Assignee: Image Metrics Ltd. (Image Metrics, Inc.)
• Original Assignee: Image Metrics Ltd. (Image Metrics, Inc.)
• Inventors: Haslam, Jane, Walker, Kevin, Edwards, Gareth
• Primary Examiner(s): Nguyen; Phu K.
• Application Number:
  US11/292,423
• Publication Number:
  US 20060126928A1
• Time in Patent Office: 788 Days
• Field of Search: 345/473, 345/474, 345/475, 345/420, 382/106, 382/286
• US Class Current: 345/473
• CPC Class Codes: G06T 13/40 of characters, e.g. humans, animals or virtual beings
• CPC Class Codes: G06T 17/20 Finite element generation, e.g. wire-frame surface description, tesselation
• CPC Class Codes: G06T 7/20 Analysis of motion motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51
• CPC Class Codes: G06V 40/171 Local features and components; Facial parts eye characteristics G06V40/18; Occluding parts, e.g. glasses; Geometrical relationships

Abstract

A system and method for cleaning a set of motion capture data, wherein the data correspond to a set of point coordinates relating to a model but may include missing or inaccurate values. A subset of the motion capture data is identified that corresponds to valid points matching locations of corresponding points of the model, and a best set of model parameters is computed in such a manner as to minimize a norm of the valid points with respect to the model. A complete set of point coordinates is then predicted based on the best set of model parameters and a subsequent set of cleaned motion capture data. Actual points are inserted where validated, and the process is iterated until a specified criterion is met.

Claims

1. A method for cleaning a set of motion capture data stored as at least one image on a digital computer, the motion capture data pertaining to an object of a specified class and the data further corresponding to a set of point coordinates relating to a model characterized by a set of model parameters, the method comprising:

a. identifying a subset of the motion capture data of each single frame in turn corresponding to valid points matching locations of corresponding points of the model, based exclusively upon distances of neighboring points in motion capture data pertaining to the single frame;

b. computing a best set of model parameters to minimize a norm of the valid points with respect to the model; and

c. predicting a complete set of predicted point coordinates based on the best set of model parameters so as to form a subsequent set of cleaned motion capture data.

2. The method according to claim 1, further comprising a step of replacing a subset of the predicted point coordinates with known valid values.

3. The method according to claim 2, further comprising using the subsequent set of cleaned motion data according to step (c) as the subset of motion capture data of step (a), and iterating steps (a), (b), and (c) until a specified criterion is met.

4. The method according to claim 3, wherein the criterion is a bound on a norm characterizing the difference between the subsequent set of cleaned motion data according to step (c) and the subset of motion capture data of step (a).

5. The method according to claim 3, wherein the criterion is a specified number of iterations.

6. The method according to claim 1, further comprising creating the model from a training set of plausible shapes and shape variations of the class of the object.

7. A method according to any of claims 1 to 6, wherein the model parameters comprise an orthonormal basis set spanning a space of model vectors.

8. A method according to any of claims 1 to 6, wherein the model comprises an average shape of a subject and a set of functions describing deviations from the average shape characterized by the model parameters.

9. A program storage device readable by a machine, tangibly embodying a program of instructions executable by the machine to perform a method for cleaning a set of motion capture data, the data corresponding to a set of point coordinates relating to a model characterized by a set of model parameters, the method comprising:

a. identifying a subset of the motion capture data of each single frame in turn corresponding to valid points matching locations of corresponding points of the model, based exclusively upon distances of neighboring points in motion capture data pertaining to the single frame;

b. computing a best set of model parameters to minimize a norm of the valid points with respect to the model; and

c. predicting a complete set of predicted point coordinates based on the best set of model parameters so as to form a subsequent set of cleaned motion capture data.

10. The program storage device of claim 9, wherein the method performed by the set of machine instructions further comprises a step of iterating steps (a) through (c).

11. A system for cleaning a set of motion capture data, the data corresponding to a set of point coordinates relating to a model characterized by a set of model parameters, the system comprising:

a. means for identifying a subset of the motion capture data of each single frame in turn corresponding to valid points matching locations of corresponding points of the model based exclusively upon distances of neighboring points in motion capture data pertaining to the single frame;

b. means for computing a best set of model parameters to minimize a norm of the valid points with respect to the model; and

c. means for predicting a complete set of predicted point coordinates based on the best set of model parameters so as to form a subsequent set of cleaned motion capture data.