摘要
视频对象的提取在序列图像的分析中起着重要作用。提出一个基于内容的多层次视频的对象提取算法,利用高斯马尔可夫模型对其进行颜色和纹理的混合特征图像分割。利用Normalize-cut准则,对其运动信息进行分析,然后进行区域聚合,即得到具有语义的视频对象。对于背景运动信息较丰富的序列图像可以取得良好的提取效果。
Extracting of semantic video object plays an important role in analysis of serial images. This paper proposes a new approach of content-based multi-hierarehy semantic video object extraction. It uses Gaussian- Markov random field model for image segmentation, which aims at combining color and texture features. The grouping process is based on the analysis of motion information of the different feature classes via normalize-cut mles. The semantic video objects are extracted by the proposed method. It has good results in videos of which background has rich motion information.
出处
《信息技术》
2008年第9期31-34,共4页
Information Technology
