作　　者： ; ;

机构地区： 华南理工大学机械与汽车工程学院

出　　处： 《计算机应用》 2009年第B06期231-232,235,共3页

摘　　要： 提出了一种基于HSI模型的彩色图像背景差法,充分考虑了色度(H)、饱和度(S)、亮度(I)三者之间的关系。根据H、S分量与I分量相互独立的特点,综合利用H分量和S分量新建背景亮度信息,根据亮度信息,采用动态阈值提取出精确的前景对象。该算法高速有效,对噪声和光线的变化有较强的鲁棒性。 This article proposed a color image background difference based on the HSI （Hue-Saturation-Intensity） model, this algorithm fully considered the relations between H, S and I. According to H and S component was independent with I component mutually, it built a new background brightness information by utilizing H and S component. According to the luminance information, the study used the dynamic threshold value to withdraw the precise prospect object. This algorithm has high speed and effectiveness, and it also has the strong robustness to the noise and the cbange of light.

关 键 词： 运动目标检测 颜色空间 背景差法 差分 动态阈值

领　　域： [自动化与计算机技术] [自动化与计算机技术]