基于高斯混和模型与Blob分析的人手定位

在基于机器视觉的手势识别研究中，手势分割与定位是关键技术。本文在分析肤色的颜色特征和人手运动特性的基础上，提出了一种在视频图像序列中实现手势分割和定位的算法。首先结合肤色检测技术和基于高斯混合模型的运动目标检测技术，获得了图像序列的初始手势区域，之后利用Blob技术实现了最终手势的分割和定位。实验结果显示，该方法具有较高的分割和定位准确度。

Locating of Human Hand Based on Gaussian Mixture Model and Blob Analysis

Gesture recognition and segmentation are key technologies of vision-based hand gesture recognition. After analyzing the skin color information features and gesture motion characteristics, we proposed an algorithm of achieve gesture segmentation and positioning from a sequence of video images. First, obtained an initial gesture area form image sequence using skin detection and moving target detection technology based on Gaussian mixture model, then achieved the ultimate gesture region based on Blob analysis. Experimental results show that this method has a high degree gesture segmentation and positioning accuracy.