基于LRP的自适应运动目标检测算法

Adaptive Moving Object Detecting Based on LRP

传统的运动目标检测算法主要基于像素值的统计模型,对于光照突变和噪声极为敏感.为此,提出了一种基于局部比率模式(LRP)的自适应运动目标检测算法.使用LRP描述视频图像序列中像素特征,通过自适应核密度估计对像素特征进行建模,提取出运动目标.实验结果表明,该算法适应光照变化,有良好的检测性能.

Traditional moving object detecting algorithms are mainly based on statistical model by pixel value, which are extremely sensitive to illumination variance and noises. To resolve this problem, a novel adaptive moving object detecting method was proposed by model pixels with its local ratio pattern （LRP）. Pixel features in video image sequences were conducted by LRP, and the moving objects were extracted through kernel density estimation. The results of experiments on L2R databases demonstrate that the proposed has excellent detection accuracy on complex scenes. adaptive algorithm