摘要
采用背景差分法检测运动目标时常常将阴影点检测成前景点,对目标分割与提取产生严重影响。为了准确提取运动目标,提出了一种基于YUV颜色空间色度畸变和一阶梯度模型进行阴影去除的方法。实验表明该算法抗干扰能力强,而且复杂度较低,易于实现实时运动图像处理。
Shadow is always regarded as foreground in detecting moving object by using background subtraction algorithm, which has bad effect on segmenting and extracting object. In order to extract object exactly, a shadow suppression method based on chrominance distortion in YUV color space and first order gradient information is proposed in this paper. Experiments result show that the algorithm is robust in noisy environment, low complexity, and easy to be implemented in real time.
出处
《微处理机》
2008年第5期116-117,121,共3页
Microprocessors
关键词
阴影去除
运动目标检测
颜色空间
一阶梯度
Shadow suppression
Moving object detection
Color space
First order gradient
