| Autor: |
Victoria Dymchuk, Albert Lekakh, Roman Puhachov, Natalia Korolyova, Dmytro Zhuikov, Victoriya Himenko |
| Rok vydání: |
2019 |
| Předmět: |
|
| Zdroj: |
2019 IEEE International Conference on Advanced Trends in Information Theory (ATIT). |
| DOI: |
10.1109/atit49449.2019.9030443 |
| Popis: |
The article substantiates the basic principles for constructing a method for recovering compressed dynamic images of a stationary background. Basic approaches for improving technologies for syntactical coding of transformants with the formation of the space of identifiers of discrete positions of a two-dimensional structural-spectral description are substantiated. An effective syntactic representation of the transformed P-type frame segments is developed on the basis of one-dimensional two-base object-position coding under conditions of flexible pair-element inequality. It is proposed to perform reconstruction of compressed dynamic images of a stationary background on the basis of: reconstruction of a binary mask matrix, reconstruction of a dynamic component, reconstruction of a matrix of signs. A technology has been built to restore the current frame in the original dynamic space without losing information based on the recovery of the current differential-presented frame and information about the base frame. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
