基于颜色均衡和椭圆模型的手势图像分割  

Elliptical Image Segmentation Based on Color Balance and Elliptical Model

作　　者：严秋锋 王红茹[1] 季鸣[1] 

机构地区：[1]江苏科技大学机械工程学院,江苏镇江212003

出　　处：《计算机仿真》2015年第4期172-175,413,共5页Computer Simulation

基　　金：江苏科技大学高级人才项目(35020902)

摘　　要：在手势图像识别系统的研究中,手势图像分割效果的好坏直接影响到手势识别系统的识别率。椭圆模型是手势图像分割中经常使用的一种肤色模型,但传统的椭圆模型易受光照的影响,对肤色的误判率较高。首先对待分割图像进行颜色均衡处理来减少光照的影响,然后对亮度信息分量设置阈值,进行分段判断,在高亮度区域扩大椭圆模型的长短轴,进一步减少光照影响。实验结果证明,上述方法比传统椭圆模型分割效果好,能够精确的分割出手势图像。Gesture segmentation directly affects the recognition rate of gesture recognition system. Elliptical model is a color image segmentation model which is often used in gesture recognition, but the traditional elliptical model is susceptible to light, the rate of false is higher. In the paper, firstly, we color carried out the image segmentation equalization to reduce the effect of light. And then, we set the threshold for luminance information component, carded out segmented judgment, and expanded the length of the axis of the ellipse model in the high brightness area, further reducing the influence of light. Experimental results show that the segmentation effect of the proposed method is better than the traditional elliptical model.

关 键 词：肤色模型 椭圆模型 颜色均衡 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]