A color based system using multiple templates was developed and implem ented for detecting human faces in color images. The algorithm consists of three image processing steps. The first step is human skin color stati...A color based system using multiple templates was developed and implem ented for detecting human faces in color images. The algorithm consists of three image processing steps. The first step is human skin color statistics. Then it separates skin regions from non-skin regions. After that, it locates the fronta l human face(s) within the skin regions. In the first step, 250 skin samples from persons of different ethnicities are used to determine the color distribution o f human skin in chromatic color space in order to get a chroma chart showing lik elihoods of skin colors. This chroma chart is used to generate, from the origina l color image, a gray scale image whose gray value at a pixel shows its likelih ood of representing the skin. The algorithm uses an adaptive thresholding proces s to achieve the optimal threshold value for dividing the gray scale image into separate skin regions from non skin regions. Finally, multiple face templates ma tching is used to determine if a given skin region represents a frontal human fa ce or not. Test of the system with more than 400 color images showed that the re sulting detection rate was 83%, which is better than most color-based face dete c tion systems. The average speed for face detection is 0.8 second/image (400×300 pixels) on a Pentium 3 (800MHz) PC.展开更多
Conventional correlation matching algorithms waste great time in invalid area search. This paper proposes a color tracking method based on correlation search area optimization on target characteristic hue decision. By...Conventional correlation matching algorithms waste great time in invalid area search. This paper proposes a color tracking method based on correlation search area optimization on target characteristic hue decision. By quantifying and reducing dimensions of HSV（ hue saturation value） color space, a one-dimensional hue space is constructed. In the space, the target characteristic hue granule set is constructed, which contains attributes such as value, area and average distance between pixels and aiming center. By using granular computing method, the similarity between target and search blocks is obtained and the invalid search areas can be removed. The color tracking experiment has proved that the algorithm can improve real time performance for conventional matching algorithms without precision lost.展开更多
文摘A color based system using multiple templates was developed and implem ented for detecting human faces in color images. The algorithm consists of three image processing steps. The first step is human skin color statistics. Then it separates skin regions from non-skin regions. After that, it locates the fronta l human face(s) within the skin regions. In the first step, 250 skin samples from persons of different ethnicities are used to determine the color distribution o f human skin in chromatic color space in order to get a chroma chart showing lik elihoods of skin colors. This chroma chart is used to generate, from the origina l color image, a gray scale image whose gray value at a pixel shows its likelih ood of representing the skin. The algorithm uses an adaptive thresholding proces s to achieve the optimal threshold value for dividing the gray scale image into separate skin regions from non skin regions. Finally, multiple face templates ma tching is used to determine if a given skin region represents a frontal human fa ce or not. Test of the system with more than 400 color images showed that the re sulting detection rate was 83%, which is better than most color-based face dete c tion systems. The average speed for face detection is 0.8 second/image (400×300 pixels) on a Pentium 3 (800MHz) PC.
文摘Conventional correlation matching algorithms waste great time in invalid area search. This paper proposes a color tracking method based on correlation search area optimization on target characteristic hue decision. By quantifying and reducing dimensions of HSV（ hue saturation value） color space, a one-dimensional hue space is constructed. In the space, the target characteristic hue granule set is constructed, which contains attributes such as value, area and average distance between pixels and aiming center. By using granular computing method, the similarity between target and search blocks is obtained and the invalid search areas can be removed. The color tracking experiment has proved that the algorithm can improve real time performance for conventional matching algorithms without precision lost.
