一种基于贝叶斯网络的视频静态语义探测新方法

传统的基于内容的视频检索是利用图像的颜色、纹理以及形状等底层特征来对视频进行检索,然而这些底层特征并没有深层次地挖掘出视频的语义内容。在用支持矢量机对图像进行分类的基础上,提出了一种基于贝叶斯网络的对视频静态语义如室内/室外进行探测的新方法,实验结果验证该方法的有效性。

A New Bayesian Network Based Approach for Video Static Semantic Detection

Traditional content-based video retrieval is using the low-level features such as color,texture and shape for video retrieval,but these low-level features doesn't mine the video semantic deeply.Based on the classification of images by support vector machine,a new Bayesian Network based approach for detection of video static semantic such as indoor/outdoor is presented in this paper and the results of experiment verify the validity of the method.