Search for a command to run...
The study addresses the development of intelligent services for the transmission of infrared (IR) imagery, with a focus on the generation and exploitation of semantic metadata that reflects the informativeness and spatial structure of image regions. Such metadata facilitates classification of object types and states and supports the selection of informative fragments within IR frames. These functionalities are particularly relevant in applications involving unmanned onboard systems, where tasks such as dynamic object tracking and autonomous navigation impose stringent requirements on the timeliness and fidelity of visual data. With the increasing complexity of operational scenarios, the demands for high information completeness – such as higher frame rates, improved resolution, and reduced distortion – place significant burdens on infocommunication subsystems. This induces a fundamental conflict between the need for low-latency transmission and the preservation of the structural integrity of IR data. To address this issue, the paper formulates a new scientific and applied objective: to enhance the quality of IR-based intelligent services through the design of a positional coding method in differential wave space. The proposed method operates by hierarchically segmenting IR images to localize structurally homogeneous regions, applying differential wave transformations to capture local contrast dynamics, and encoding positional information relative to wavelet-domain variations. This positional representation enables spectral-differential group coding, which effectively reduces bit volume while preserving spatial-semantic features. An experimental evaluation using the Open Turbulent Image Set (OTIS) – consisting of PNG-encoded IR images with varying levels of informativeness – demonstrated an average bit volume reduction of approximately 37%. Comparative analysis with JPEG2000 encoding of 16-bit images revealed a 25% gain in compression ratio while maintaining equivalent PSNR values.
Published in: Informatyka Automatyka Pomiary w Gospodarce i Ochronie Środowiska
Volume 16, Issue 1, pp. 138-146
DOI: 10.35784/iapgos.7776