摘要
为了校正由于光源颜色变化而引起的图像色彩的偏差,在RGB颜色空间基于GrayW orld假设,采用颜色平衡方法对输入的彩色图像进行颜色校正;为了有效地检测彩色图像中的肤色区域,采用了新颜色空间YCgCr,并在该颜色空间建立了联合亮度信息和Cg-Cr色度查找表的肤色模型。大量实验结果表明,该模型具有较强的光照适应性,可应用于复杂环境下的彩色图像中的人体肤色区域检测。
To correct the image color bias produced by the illuminant, a color balance method was introduced in the RGB color space based on the Gray World hypothesis. To detect skin-color regions in the image efficiently, a new color space YCgCr was adopted, and a skin-color model of combining luminance with Cg-Cr chroma lookup table was constructed in the new color space. Experimental results demonstrate that this model has more robust illumination adaption, it can be used to detect human body skin colors in complex scene images.
出处
《计算机应用研究》
CSCD
北大核心
2008年第2期467-468,472,共3页
Application Research of Computers
基金
国家自然科学基金资助项目(60072034)
