Adaptive orthogonal transformation of video image method

In the article the important research and technology task of development and research methods and video data compression algorithms. In many social spheres of human activity means of objective remote control and monitoring are applied. Such means on the basis of analysis of data received in real time. The same social spheres the following aero video monitoring of traffic situation in big megalopolis, environmental resources, prevention and operational of emergency recovery, etc.

Generally, in such video monitoring systems video data image detecting data are transmitted to point data analysis by broadband heterogeneous networks. The heterogeneous is caused by resort employment of combination different communication technologies. This and broadband networking technology, which are used, for example, information transmission is provided from an unmanned aerial vehicles to a ground control station. Also batch high-speed communication systems, which are provided information transmission for subscribers of the terrestrial infrastructure. Accordingly an important aspect of similar systems is providing its high-capacity, minimum data loss during transmission and also providing minimum data garbling. In such a way the development task of adoptive data compression methods is the actual research and technology task. As well as quality requirement such video data are placed increased demands on video data quality from different user application. Besides, there is a need to parameter optimization of communication link by capacity and possible timing delay during compressed video data transfer.

The article includes synthesis of adaptive algorithm for video data compressed on the basis of discrete cosine transformation methods and wavelet transform. Analytical techniques is shown that proposed adaptive procedures video data compressed allow to reduce structural and parametric by communicated information about 25—50fold with no loss in quality. On the basis of existing methods and algorithms video data compressed by analytical approach is shown that appliance of existing methods is limited, in consequence of low data-compression ratio (2—4 fold), which methods and algorithms provide. Application of the offered procedures allows reducing such limitation. As received compression ratios are large in the same procedures in comparison with known methods. From the analysis of attained result by computational modeling research and technology results are received about the adoptive orthogonal transformation is most resistant to link errors recovery video data by the proposed method has not artifacts as blocks, which are typical for discrete cosine transform and artifacts as boundary detection which are typical for wavelet transform. In such away the developed method of adaptive orthogonal transformation has a number of adaptive before known methods by orthogonal transformation.