The work analyzes technical and experimental research in the field of fire monitoring. The authors suggest the algorithms of fire detection, the identification of the seats of fire, the modeling, the determination of the spreading and the following of the fire by satellite images processing. The methodology is implemented by the creation of the intelligent telecommunication structures on the base of neuroprocessor systems NM640X capable to learn in the conditions of uncertainty.
The article is devoted to reducing the risk of decision-making on the natural and manmade critical emergency prevention by means of specialized processing of remote earth sensing data on the basis of neural fuzzy inference networks. The fire monitoring and predicting algorithm based on the spectrum processing of the satellite images is suggested. The operability testing of the suggested technology was conducted on the base of NM640X neuroprocessor.
The work analyzes technical and experimental research in the field of fire monitoring. It proposes a technique of using cyber-physical systems for monitoring natural and technogenic emergencies, providing information and algorithmic support for a cyber-physical system of forest-fire monitoring. It also provides hierarchical structures of components for a cyber-physical system of forest-fire monitoring. The methodology is implemented as development of intelligent telecommunications structures on the base of the MB 7707 neuroprocessor systems that can undergo training in conditions of uncertainty. The work is accomplished within the framework of the Russian Foundation for Basic Research (RFBR) grant 14-07-00261A.
The paper suggests a generic model for compilation and definition of explicit parallelism in embedded computing systems (ECS), based on clustering. It also describes a rule-oriented production model of an expert system for analyzing implicit parallelism that employs a criteria-based fuzzy-set approach, the number of such criteria equal to the number of ECS modules, and also on the microprogram's memory capacity, the system's performance, etc.
The article regards digital images as mathematical objects and uses a co-variation matrix of element brightness values to analyze an autocorrelation matrix of images – a dispersion matrix, to define all significant properties of images with the help of mathematical statistics and matrix computing. We suggest theoretic and methodological generalizations in the field of aerospace photo processing, basing on choice of the most advantageous variant of constructing basic functions among various Vilenkin-Chrestenson functions (VCF) in a non-trigonometrical minimally available construction form as Walsh functions with an original ‘carcass’ of Rademacher functions. The methodology of this research is based on a set of theorems that are different from corresponding theorems of classical spectral analysis [5]. The conclusions of the theorems allow constructing effective algorithms of filtration and correlation analysis of aerospace images when noises and distortions appear.
World experience shows the effectiveness of continuous monitoring of the fire hazardous areas to carry out management for decisions to prevent emergencies [1]. Identification of fires with the help of the images obtained by remote sensing of the earth surface, is essential to the economy, since timely identification of fire gives a real opportunity to localize and eliminate it in a small area, thus preventing its further spread. It helps to reduce material costs for fire-fighting and the amount of recovery work.