US6975750B2
Expired / Lapsed
US
⏱️ 0.0 years remaining
System and method for face recognition using synthesized training images
Assignee: Zhigu Holdings Ltd
•
Filing: 2000-12-01
•
Granted: 2005-12-13
•
Expires: 2020-12-01
Abstract
Abstract A system and method that includes a virtual human face generation technique which synthesizes images of a human face at a variety of poses. This is preferably accomplished using just a frontal and profile image of a specific subject. An automatic deformation technique is used to align the features of a generic 3-D graphic face model with the corresponding features of these pre-provided images of the subject. Specifically, a generic frontal face model is aligned with the frontal image and a generic profile face model is aligned with the profile image. The deformation procedure results in a single 3-D face model of the specific human face. It precisely reflects the geometric features of the specific subject. After that, subdivision spline surface construction and multi-direction texture mapping techniques are used to smooth the model and endow photometric detail to the specific 3-D geometric face model. This smoothed and texturized specific 3-D face model is then used to generate 2-D images of the subject at a variety of face poses. These synthesized face images can be used to build a set of training images that may be used to train a recognition classifier.
Inventors
- 👤 Jie Yan
- 👤 Hong-Jiang Zhang
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 10:19
CPC Classifications
G; G06; G06T; G06T17/00; G06F; G06F18/00; G06F18/20; G06F18/21; G06F18/214; G06V; G06V40/00; G06V40/10; G06V40/16; G06V40/172
Claims Analysis
Total Claims: 25
Claim 1
Independent Claim
98 words
A computer-implemented process for face recognition, comprising using a computer to perform the following process actions: inputting an image of a face of a subject sought to be recognized having a particular face pose; inputting a generic three dimensional face model; creating a specific three dimensional face model of the specific subject sought to be recognized by deforming the generic face model to conform to the shape of the face depicted in the input image; synthesizing various face pose images using the specific 3-D face model; and employing the synthesized images as training images to train a recognizer.
Claim 2
Dependent Claim
98 words
A computer-implemented process for face recognition, comprising using a computer to perform the following process actions: inputting an image of a face of a subject sought to be recognized having a particular face pose; inputting a generic three dimensional face model; creating a specific three dimensional face model of the specific subject sought to be recognized by deforming the generic face model to conform to the shape of the face depicted in the input image; synthesizing various face pose images using the specific 3-D face model; and employing the synthesized images as training images to train a recognizer.
Claim 3
Dependent Claim
24 words
The computer-implemented process of claim 1 further comprising the process action of using a spline surface construction technique to smooth the specific face model.
Claim 4
Dependent Claim
26 words
The computer-implemented process of claim 2 further comprising the process action of using a texture mapping technique to endow textural detail to the smoothed face model.
Claim 5
Dependent Claim
29 words
The computer-implemented process of claim 1 wherein the process action for inputting an image of a face comprises inputting one frontal view and one profile view of the face.
Claim 6
Dependent Claim
37 words
The computer-implemented process of claim 1 wherein the process action for inputting an image comprises inputting at least two views of a face each having poses varying in orientation from the others by at least 15 degrees.
Claim 7
Dependent Claim
39 words
The computer-implemented process of claim 1 wherein the process action for inputting a generic three dimensional face model comprises employing a polygon model that depicts the surfaces of the face as a series of vertices defining a facial mesh.
Claim 8
Dependent Claim
44 words
The computer-implemented process of claim 6 wherein the process action for creating a specific three dimensional face model of the subject comprising using a deformation technique to create a specific three dimensional model of the subject from the generic model and the input images.
Claim 9
Dependent Claim
61 words
The computer-implemented process of claim 7 wherein the process action of using a deformation technique to create a specific three dimensional model further comprises the process actions of: extracting feature point sets from input frontal and profile images of the subject; and modifying the generic face model to the specific face model by comparison and mapping between the feature point sets.
Claim 10
Dependent Claim
48 words
The computer-implemented process of claim 4 wherein symmetry of the face is assumed, such that for the purposes of said creating of a specific 3-D face model it is assumed the side of the face opposite that depicted in the profile image has the same shape and coloring.
Claim 11
Dependent Claim
38 words
The computer-implemented process of claim 1 wherein two or more images are used to create the model of a specific face, further comprising a process action of refining the specific model of the face by using additional images.
Claim 12
Dependent Claim
21 words
The computer-implemented process of claim 2 wherein the facet-based representation of the specific face is replaced with a spline surface representation.
Claim 13
Dependent Claim
41 words
The computer-implemented process of claim 3 wherein the process action to endow textural detail to the face model further comprises the process action of using the input images to assign color intensity to each pixel of the three dimensional face model.
Claim 14
Dependent Claim
21 words
The computer-implemented process of claim 12 wherein a Bézier patch technique is employed to add texture detail to a face model.
Claim 15
Dependent Claim
47 words
The computer-implemented process of claim 13 wherein the process action for assigning color intensity further comprises the process action of: for each Bézier surface patch of the face surface determining a corresponding texture patch by mapping the boundary curve of the Bézier patch to the face image.
Claim 16
Dependent Claim
80 words
The computer-implemented process of claim 13 wherein both frontal and profile images are employed and wherein the face image chosen to provide the texture information depends on the preferred direction of the Bézier patch; such that the frontal image is used to map when the angle between the direction and the Y-Z plane is less than 30 degrees and wherein a profile image is used when the angle between the direction and the Y-Z plane is greater than 30 degrees.
Claim 17
Dependent Claim
43 words
The computer-implemented process of claim 1 wherein face pose images are synthesized for every in-plane rotation of plus or minus 10 to 15 degrees and every out-of-plane rotation of plus or minus 15-20 degrees, with increments of about 3-7 degrees within a group.
Claim 18
Dependent Claim
36 words
The computer-implemented process of claim 1 wherein the illumination is varied to produce the synthetic images at varied illuminations by taking the generated images and synthetically varying the illumination to produce each image at various illuminations.
Claim 19
Independent Claim
123 words
A system for generating synthesized face images, the system comprising: at least one camera positioned so capture at least one image of a subject for whom images are to be synthesized; a general purpose computing device; and a computer program comprising program modules executable by the computing device, wherein the computing device is directed by the program modules of the computer program to, input at least one image of a face of a subject from the at least one camera; input a generic face model; create a specific face model of the subject by deforming the generic face model to conform to the shape of the face depicted in the input image; and synthesize various face poses using the specific 3-D face model.
Claim 20
Independent Claim
123 words
A system for generating synthesized face images, the system comprising: at least one camera positioned so capture at least one image of a subject for whom images are to be synthesized; a general purpose computing device; and a computer program comprising program modules executable by the computing device, wherein the computing device is directed by the program modules of the computer program to, input at least one image of a face of a subject from the at least one camera; input a generic face model; create a specific face model of the subject by deforming the generic face model to conform to the shape of the face depicted in the input image; and synthesize various face poses using the specific 3-D face model.
Claim 21
Dependent Claim
27 words
The system of claim 18 , further comprising a computer module to, employ the synthesized images as training images by inputting the synthesized images into a recognizer.
Claim 22
Dependent Claim
23 words
The system of claim 18 , further comprising a computer module to, use a spline surface technique to smooth the specific face model.
Claim 23
Dependent Claim
26 words
The system of claim 20 , further comprising a computer module to, use a texture mapping technique to endow textural detail to the smoothed face model.
Claim 24
Dependent Claim
117 words
A computer-readable medium having computer-executable instructions for generating synthesized images of a face, said computer-executable instructions comprising: inputting an image of a face of a subject having a particular face pose; inputting a generic three dimensional face model; creating a specific three dimensional face model of the specific subject by deforming the generic face model to conform to the shape of the face depicted in the input image; using a spline surface construction technique to smooth the specific face model; using a texture mapping technique to endow textural detail to the smoothed face model; synthesizing various face pose images using the specific 3-D face model; and employing the synthesized images as training images to train a recognizer.
Claim 25
Dependent Claim
117 words
A computer-readable medium having computer-executable instructions for generating synthesized images of a face, said computer-executable instructions comprising: inputting an image of a face of a subject having a particular face pose; inputting a generic three dimensional face model; creating a specific three dimensional face model of the specific subject by deforming the generic face model to conform to the shape of the face depicted in the input image; using a spline surface construction technique to smooth the specific face model; using a texture mapping technique to endow textural detail to the smoothed face model; synthesizing various face pose images using the specific 3-D face model; and employing the synthesized images as training images to train a recognizer.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2008-12-12 | 2009-06-13 | 2009-12-13 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2012-12-12 | 2013-06-13 | 2013-12-13 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2016-12-12 | 2017-06-13 | 2017-12-13 | $7700.00 | Window Closed / Lapsed |