中文说明：本系统设计并实现了一种基于色度空间的最邻近n帧改进算法，以提高传统混合高斯模型的目标检测效果。当光照发生突然变化时，利用像素点的色度信息代替RGB值，有效避免亮度突变的干扰；为了反映背景的动态变化，检测系统使用各高斯分布在最近n帧与像素点测量值的匹配次数的指数函数实现权重值的更新。
English Description：
This system designs and implements an improved algorithm of nearest n frames based on chroma space to improve the target detection effect of traditional Gaussian mixture model. When the illumination changes suddenly, the pixel chroma information is used to replace the RGB value to avoid the interference of brightness mutation; in order to reflect the dynamic changes of the background, the detection system uses the exponential function of the matching times of each Gaussian distribution in the nearest n frames and the measured value of the pixel to update the weight value.