The article provides a general description of the approach to building a code representation of video stream fragments, namely, a DCT transformer. Such code representation is a set of codograms of individual bit planes of transformants. In turn, a codogram of each bit plane is constructed using the detection of the lengths of a series of binary elements within each of them. Detection of a series of binary elements within the proposed method is in the direction of the rows. The resulting bitmap codes are formed on the basis of non-equilibrium position codes. Unlike traditional coding techniques that aim at reducing probabilistic statistical redundancy, in this case, the structural redundancy of the bit description of video stream frame transforms is eliminated. In this case, the code description of each of the transformants is formed separately. In other words, within the code structures, the transformants of each of the codograms of its bit planes are independent of each other. That is, with the exception of one or more codograms that carry a description of one or another number of bit planes, no transformants occur on the receiving side of the destruction of the entire code description. That is, the generated code description allows you to repair the transformant on the receiving side with some errors. It turns out that such manipulation of the number of codograms that were initially used to describe the transform of the video stream frame allows to change the number of bits required for its code description. Thus, in this way, there is actually an additional reduction in the information intensity of the transmitted data, in addition to the reduction that is achieved by actually translating the transformant description to the appearance of the position code. In turn, this makes it possible to use the encoding method in question as a basis for video intensity management technology. The developed method does not require significant computing power for its effective functioning, since the construction of the bit description code requires less hardware resources by reducing the dynamic range of computation compared to component coding.