System and method for adaptive skin tone detection

A system for detecting human skin tone in an image, said system comprising:

• a camera configured to capture one or more images;
• an image processing module configured to detect the presence of skin tone in said image, said image processing module comprising:
• a contrast normalization module configured to receive said image and enhance contrast of said image and generate a contrast-enhanced image;
• and
• a skin segmentation module configured to convert said contrast-enhanced image into multiple color spaces and process each pixel of said image to identify values of each pixel corresponding to each of said multiple color spaces, said skin segmentation module further configured to classify said image into skin and non-skin regions based on a comparison of said identified values of each pixel with corresponding values of skin tone boundaries of a skin database;
• wherein said skin segmentation module is configured to generate output data indicative of said identified skin regions and non-skin regions of said image;
• and
• a media device configured to receive said output data from said image processing module for presentation or further processing.

At least one computer accessible medium storing instructions which, when executed by a machine, cause the machine to perform operations for detecting human skin tone in an image, said operations comprising:

• receiving an image;
• enhancing contrast of said image and generating a contrast-enhanced image;
• converting said contrast-enhanced image into multiple color spaces;
• processing each pixel of said image to identify values for each color space;
• generating histograms for one or more values of each pixel for each color space and comparing said histograms with skin tone boundaries having known skin tone values to identify pixels containing values within lower and upper limit thresholds of said skin tone boundaries;
• classifying said image into skin regions and non-skin regions based on said comparison of said histograms with said skin tone boundaries;
• and
• generating output data indicative of said identified skin regions and non-skin regions of said image.

A method for detecting human skin tone in an image, said method comprising:

• receiving and enhancing, by a contrast normalization module, contrast of an image and generating a contrast-enhanced image;
• converting, by a skin segmentation module, said contrast-enhanced image into multiple color spaces and processing each pixel of said image to identify values for each color space;
• generating, by said skin segmentation module, histograms for one or more values of each pixel for each color space and comparing said histograms with skin tone boundaries, of a skin tone database, having known skin tone values to identify pixels containing values within lower and upper limit thresholds of said skin tone boundaries;
• classifying, by said skin segmentation module, said image into skin regions and non-skin regions based on said comparison of said histograms with said skin tone boundaries;
• and
• generating output data indicative of said identified skin regions and non-skin regions of said image.