Image correction method, image correction apparatus and video system

Patent metadata

Field	Value
Publication number	US-2016358318-A1
Application number	US-201514891943-A
Country	US
Kind code	A1
Filing date	May 18, 2015
Priority date	Jan 29, 2015
Publication date	Dec 8, 2016
Grant date	—

Abstract

The present invention provides an image correction method, an image correction apparatus and a video system. The image correction method comprises: obtaining an actual face image; obtaining a target sample image matching with the actual face image; and correcting each eye area of the actual face image according to the target sample image such that orientation of each eye area of the corrected actual face image coincides with orientation of a corresponding eye area of the target sample image. The present invention may enable pupil portions of eye areas in a received face image to look toward a camera, thereby realizing an equal communication and improving sensory experience.

First claim

1 . An image correction method, comprising:
obtaining an actual face image; obtaining a target sample image matching with the actual face image; and correcting each eye area of the actual face image according to the target sample image such that orientation of each eye area of the corrected actual face image coincides with orientation of a corresponding eye area of the target sample image.

2 . The image correction method according to claim 1 , wherein the step of obtaining a target sample image matching with the actual face image comprises:
obtaining feature information of the actual face image; and calculating matching degree between the feature information of the actual face image and feature information of each preset sample image in a sample library respectively, and setting, when the matching degree between a preset sample image and the actual face image is larger than a preset value, the preset sample image of which the matching degree with the actual face image is larger than the preset value as the target sample image.

3 . The image correction method according to claim 2 , wherein the feature information comprises any one or more of the following: a center coordinate of each eye area, an aspect ratio of each eye area, a diameter of an iris in each eye area, a ratio of a distance between a center of a left eye area and the nose to a distance between a center of a right eye area and the nose, and a ratio of a distance between a center of a left eye area and the mouth to a distance between a center of a right eye area and the mouth.

4 . The image correction method according to claim 2 , further comprising the following step between the step of obtaining a target sample image matching with the actual face image and the step of correcting each eye area of the actual face image according to the target sample image such that orientation of each eye area of the corrected actual face image coincides with orientation of a corresponding eye area of the target sample image:
obtaining a center coordinate of a pupil portion of each eye area in the actual face image and a center coordinate of a pupil portion of each eye area in the target sample image; and calculating a distance between the center coordinate of the pupil portion of each eye area in the actual face image and the center coordinate of the pupil portion of the corresponding eye area in the target sample image, and performing, when the distance is large than a predefined value, the step of correcting each eye area of the actual face image according to the target sample image such that orientation of each eye area of the corrected actual face image coincides with orientation of the corresponding eye area of the target sample image.

5 . The image correction method according to claim 1 , wherein the image correction method further comprises the following steps performed before the step of obtaining an actual face image:
collecting a plurality of face images as preset sample images, orientation of each eye area of the human being collected in each preset sample image directly facing a camera; and obtaining and storing feature information of each preset sample image to form the sample library.

6 . The image correction method according to claim 5 , wherein the image correction method further comprises the following step performed after the step of obtaining and storing feature information of each preset sample image:
classifying the preset sample images into a plurality of classes according to a predefined rule; and the step of calculating matching degree between the feature information of the actual face image and feature information of each preset sample image in a sample library respectively comprises: determining a class of preset sample images corresponding to the actual face image; and calculating matching degree between the feature information of the actual face image and the feature information of each preset sample image within the class of preset sample images corresponding to the actual face image.

7 . The image correction method according to claim 6 , wherein the predefined rule comprises a maximum likelihood classification method.

8 . The image correction method according to claim 1 , wherein the image correction method further comprises the following steps performed before the step of obtaining an actual face image:
establishing a coordinate axis within a visual field range of the camera; setting a predefined coordinate range within the visual field range, the predefined coordinate range being greater than a distance between center coordinates of two eye areas of the target sample image; and setting, when a center coordinate value of any one of eye areas of the actual face image exceeds the predefined coordinate range, a boundary value of the predefined coordinate range as an actual coordinate value of the center of the eye area in the actual face image which exceeds the predefined coordinate range.

9 . The image correction method according to claim 1 , wherein the step of correcting each eye area of the actual face image according to the target sample image such that orientation of each eye area of the corrected actual face image coincides with orientation of a corresponding eye area of the target sample image comprises:
obtaining information of feature points of each eye area in the actual face image and information of target feature points, in the corresponding target sample image, corresponding to the feature points of each eye area in the actual face image; calculating mapping relationship between the feature points of each eye area in the actual face image and the corresponding target feature points; and calculating a corrected position of each feature point of each eye area in the actual face image according to the mapping relationship.

10 . An image correction apparatus, comprising:
an actual image obtaining module for obtaining an actual face image; a target sample image obtaining module for obtaining a target sample image matching with the actual face image; and a processing module for correcting each eye area of the actual face image according to the target sample image such that orientation of each eye area of the corrected actual face image coincides with orientation of a corresponding eye area of the target sample image.

11 . The image correction apparatus according to claim 10 , wherein the target sample image obtaining module comprises:
a feature information obtaining unit for obtaining feature information of the actual face image; and a matching degree calculating unit for calculating matching degree between the feature information of the actual face image and feature information of each preset sample image in a sample library respectively, and setting, when the matching degree between a preset sample image and the actual face image is larger than a preset value, the preset sample image of which the matching degree with the actual face image is larger than the preset value as the target sample image.

12 . The image correction apparatus according to claim 11 , wherein the feature information comprises any one or more of the following: a center coordinate of each eye area, an aspect ratio of each eye area, a diameter of an iris in each eye area, a ratio of a distance between a center of a left eye area and the nose to a distance between a center of a right eye area and the nose, and a ratio of a distance between a center of a left eye area and the mouth to a distance between a center of a right eye area and the mouth.

13 . The image correction apparatus according to claim 10 , wherein the image correction apparatus further comprises:
a coordina

Assignees

• Boe Technology Group Co Ltd

Inventors

• Zhang Lijie
• Wu Naifu

Classifications

• G06F18/2415

  based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate · CPC title

• G06V10/993

  Evaluation of the quality of the acquired pattern · CPC title

• G06K9/00281

  Physics · mapped topic

• G06K9/00288

  Physics · mapped topic

• H04N7/15

  Conference systems · CPC title