基于RS的关键帧提取仿真研究

摘要

研究视频图像中关键图像提取问题,视频图像传输采集效率低,且易造成资源浪费.针对传统的视频图像关键帧技术,都是以图像中的关键特征为提取依据的.当关键图像中特征不明显或者与非关键图像特征重复时,由于特征的关键帧图像提取就会发生错误,导致算法错误率和漏检率高.为解决上述问题,提出一种基RS理论的关键帧提取算法,首先提取图像DCT系数,利用RS理论的属性约简产生信息系统的核,对应到视频即为关键帧,避免了传统方法对图像关键特征的依赖.实验结果表明,算法可以提高关键帧提取的准确性和提取效率,为快速提取提供了依据.%Study key images extraction in video images. In traditional video image technology, key frames are based on the key characteristics of images. When the key image characteristics are not obvious or repetitive with non -critical image features, the method based on feature extraction may have errors, and the error rate and miss rate are high. The paper proposed a RS theory based key frame extraction method. Firstly, the algorithm extracted image DCT coefficient, used the attribute reduction of the RS theory to produce the kernel of information system, which was the key frames when corresponded to videos. The method can avoid depending key features as the traditional method does. Experimental results show that the algorithm and improve the accuracy of key frame extraction and achieve good effect.