中文摘要：

复杂环境下如何进行鲁棒的运动目标检测是计算机视觉领域热门研究课题。本文提出了一种新的码书和高斯局部二值模式（GLBP）的纹理描述的运动物体检测方法,在线学习构建码书纹理背景模型。首先用码书以类似聚类的方式构建每个像素的码书模型,根据码字的颜色和亮度相似性,将背景像素分布用聚类码字的形式表示出来,同时在模型初始化和运动检测阶段不断更新码字以反映背景变化。然后用单高斯模型来学习背景像素变化的概率,生成GLBP纹理算子,同时在线更新GLBP反映图像空间纹理信息变化。最后融合三个特征将当前帧分割为前景背景两部分。通过实验视频表明本方法在实际视频中取得了较好的鲁棒的效果。

英文摘要：

It is challenging issues to extract the moving foreground objects from background robustly in visual surveillance system. In this paper,we present a novel texture-based cluster-like algorithm to detect motion with codebook and Gaussian local binary patterns（GLBP）,which may get texture background model on-line.Firstly,a codebook model like pixel cluster is constructed.Distribution of background pixels is presented by pixel cluster using computing the color and brightness distortion between codebook and current pixel.Our algorithm updates the codebook model both in initial step and detection step to deal with changes of background pixels. A single Gaussian model of pixel-wise is used to build the pixel＇ s value change model on-line.Gaussian local binary patterns background model is constructed on-line by applying the correlation and texture of spatially proximal pixels.Finally current image is segmented into two parts,foreground and background by fusing the three features;codebook model,single Gaussian background model and Gaussian local binary patterns.Experiments show that our proposed algorithm achieves robust performance in natural videos.