Automatic correction and enhancement of facial images
AssigneeLuxand, Inc.
InventorAlexey Konoplev
Life of this patent
Estimating the term…
Abstract
A system, method and computer program product for correction and enhancement of digital images containing portraits or images of human faces by automatically detecting imperfections in an original facial image and correcting them in order to enhance the original image quality. The imperfections can be various skin blemishes, birth marks, pimples, freckles, wrinkles, etc. The facial images are detected and the imperfections are recognized. Then, the imperfections are automatically corrected by selecting a most suitable skin color using a histogram of distribution of color values on a face. A white balance and appropriate colors are set for an image. The corrected image is combined with the original image for preservations of details.
Claims13
A method for enhancing a facial image, the method comprising:
- (a) receiving an original image;
- (b) converting the original image into a device-independent color model;
- (c) processing the converted image through a skin recognition filter;
- (d) detecting faces in the converted image;
- (e) processing the converted image using wrinkle reducing parameters,
- wherein an initial filtering radius and a final filtering radius are set, and repeating steps (c)-(e) while the initial filtering radius is increased until the initial filtering radius reaches the final filtering radius;
- (f) reducing blemishes in the converted image;
- (g) calculating at least one chromaticity distribution in a facial area of the converted image;
- and
- (h) analyzing the chromaticity distribution and modifying the facial areas of the converted image based on the distribution analysis.
A system for enhancing a facial image, the system comprising:
- a processor;
- a memory coupled to the processor;
- a computer program logic stored in the memory and executed on the processor, the computer program logic for implementing:
- (a) receiving an original image;
- (b) converting the original image into a device-independent color model;
- (c) processing the converted image through a skin recognition filter;
- (d) detecting faces in the converted image;
- (e) processing the converted image using wrinkle reducing parameters,
- wherein an initial filtering radius and a final filtering radius are set, and repeating steps (c)-(e) while the initial filtering radius is increased until the initial filtering radius reaches the final filtering radius;
- (f) processing the converted image using blemish reducing parameters;
- (g) calculating at least one chromaticity distribution in a facial area of the converted image;
- and
- (h) analyzing the chromaticity distribution and modifying the facial areas of the converted image based on the distribution analysis.
Dependencies are inferred from the claim text.