S. V. Kuleshov

Since the advent of animation and cinema as a way of presenting visual information "stream" is understood as a sequence of still images, successive equal intervals of time. Development of technologies of digital representation of "moving pictures" need to expand the definition video, taking into account emerging mechanisms of its submission and processing. Feature of the video compared to a set of unrelated images is the presence of corellation dependencies between adjacent frames. Such dependence is the reason of redundancy in the representation of video as a sequence of separate images, and redundancies between adjacent frames are usually many times greater than the redundancy of representation within the frame. Analyzing the line direction in a spatial matrix for the video stream can be judged on the characteristics of compression algorithms. In this paper we propose a method for estimating interframe (temporal) redundancy on the basis of determining the number of repeating pixel values of the current frame in each subsequent frame. The results obtained allow to analyze the images uncompressed video stream to make assumptions about the used codec compression type and parameters of its level of compression. In the case of codecs of MPEG images show the characteristic number and position of static blocks that the encoder does not include the flow in the B-and P-frames. The above parameters can improve the efficiency of work with video streams, compressed by MPEG codecs when performing the modification and re-packaging.