研制了一种基于运动目标检测技术的交互式图片浏览系统.首先使用高斯混合模型对背景进行建模,提出把传统的RGB颜色空间转化为YCbCr颜色空间的改进方法,以提高检测算法对光照变化的鲁棒性,实现对前景移动目标的有效检测,进而分离出目标对象,同时结合肤色检测技术和Canny边缘检测算法来实现对移动目标的定位.最后对实际开发的系统进行功能测试验证,取得了良好的预期效果.
Proposed in this paper is an interactive photo-browsing system based on the moving target detection. In this system, the Gaussian mixture model is used to construct a background model and the traditional RGB color space is converted into YCbCr color space to improve the robustness of the system to illumination varia- tion. Thus, the moving foreground target is effectively detected and the target object is further separated. More- over, in order to locate a moving object, the skin segmentation technology is combined with the Canny edge de- tection algorithm. The developed system is finally tested by experiments, with satisfactory results being ob- tained.