彩色人体图像的二值化方法

Binarization approach of color human images

为了快速有效的对彩色人体图像进行二值化处理,提出了一种新的自适应二值化处理方法。该方法根据人体肤色特征,建立人体肤色模型;根据该模型,分割出肤色和非肤色,再直接转换成二值图像。由于人体肤色受到不同光照和环境的影响,人体彩色图像中的人体肤色值会有偏移。为了改善肤色模型适应不同环境和光照,在肤色分割过程中,采用了自适应Gamma校正方法。最后,给出了该方法的实验结果。实验结果表明,提出的彩色人体图像的二值化方法快速且有效。

To solve binarization processing for color human images quickly and efficiently, a new adaptive binarization approach is proposed.Firstly, the body skin color model is built according to the features of body skin color, then according to this model, skin color and non-skin color is segmented and converted into binary images directly.Because the human skin color are exposed to different illumination and environmental, the body color in the images will be offset the value of skin color.In the skin color segmentation, adaptive Gamma correction is adopted to improve the skin color model and adapt to different environment and illumination.Finally, the experimental results are described.The experimental results show the binarization approach of color human images is rapid and effective.