Method and system of person identification by facial image
First Claim
1. A computer-based method for identification of a person by a facial image, the method comprising:
- localizing the facial image;
detecting coordinates of eye pupil centers in the facial image;
responsive to detecting the coordinates of eye pupil centers, deriving a normalized facial image based on the facial image;
splitting the normalized facial image into rectangular blocks;
describing the rectangular blocks of the normalized facial image by a DCT (discrete cosine transform);
calculating distances between the normalized facial image and existing templates, the distances being weighted sums of shortest distances between the rectangular blocks of the normalized facial image and rectangular blocks of the existing templates, wherein one or more most different blocks are filtered from the weighted sum; and
recognizing the normalized facial image based on the distances calculated.
5 Assignments
0 Petitions
Accused Products
Abstract
A method to recognize a facial image is described. An input facial image is normalized by scaling and rotation angle using methods of eye pupil centers detection. The input facial image is further normalized by lighting intensity. Template images are obtained either by the processing of certain images taken from different face positions, or by a preliminary reconstruction of a 3D face model based on stereo-pair images. Using the 3D model template, facial images are generated at different rotation angles. Distances between the input facial image and the template image are calculated from the Discrete Cosine Transformation (DCT) features, defined by overlapped blocks of these images. The facial image is recognized based on these distances.
8 Citations
28 Claims
-
1. A computer-based method for identification of a person by a facial image, the method comprising:
-
localizing the facial image; detecting coordinates of eye pupil centers in the facial image; responsive to detecting the coordinates of eye pupil centers, deriving a normalized facial image based on the facial image; splitting the normalized facial image into rectangular blocks; describing the rectangular blocks of the normalized facial image by a DCT (discrete cosine transform); calculating distances between the normalized facial image and existing templates, the distances being weighted sums of shortest distances between the rectangular blocks of the normalized facial image and rectangular blocks of the existing templates, wherein one or more most different blocks are filtered from the weighted sum; and recognizing the normalized facial image based on the distances calculated. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
-
-
17. A non-transitory machine-readable storage medium storing executable computer program instructions for identification of a person by a facial image, the instructions performing steps comprising:
-
localizing the facial image; detecting coordinates of eye pupil centers in the facial image; responsive to detecting the coordinates of the eye pupil centers, deriving a normalized facial image based on the facial image; splitting the normalized facial image into rectangular blocks; describing the rectangular blocks of the normalized facial image by a DCT (discrete cosine transform); calculating distances between the normalized facial image and existing templates, the distances being weighted sums of shortest distances between the rectangular blocks of the normalized facial image and rectangular blocks of the existing templates, wherein one or more most different blocks are filtered from the weighted sum; and recognizing the normalized facial image based on the distances calculated. - View Dependent Claims (18, 19, 20, 21, 22, 27)
-
-
23. A computer-based system for identification of a person by a facial image, the system comprising:
-
means for localizing the facial image; means for detecting coordinates of eye pupil centers in the facial image; means for deriving a normalized facial image based on the facial image responsive to determining the coordinates of eye pupil centers; means for splitting the normalized facial image into rectangular blocks; means for describing the rectangular blocks of the normalized facial image by a DCT (discrete cosine transform); means for calculating distances between the normalized facial image and existing templates, the distances being weighted sums of shortest distances between the rectangular blocks of the normalized facial image and rectangular blocks of the existing templates, wherein one or more most different blocks are filtered from the weighted sum; and means for recognizing the normalized facial image based on the distances calculated. - View Dependent Claims (24, 25, 26, 28)
-
Specification