ARTIFICIAL NEURAL NETWORKS APPLICATION IN THERMOGRAPHY Koštialová Jančíková Z. 1 , *Koštial P. 1 , Zimný O. 1 , Heger M. 1 , Špička I. 1,2 *Author for correspondence Department of Automation and Computing in Metallurgy 1 VŠB – Technical University of Ostrava, 2 University of Business and Law Czech Republic ABSTRACT In this paper, we present an application of artificial neural network (ANN) analysis in the thermovision identification of the studied thermal fields. Precise thermal field identification plays an important role in distinguished technological processes, for instance in metallurgy. Our efforts were focused in this direction. Thermovision outputs are usually thermograms with a form of a quasi-coloured imaging record of an observed temperature field. A thermogram is usually registered and presented in a form of an electronic or printed image. The character of such a document is informational only, and real temperature values are difficult to detect. The exploitation of neural networks is advantageous, if it is necessary to express complex mutual relations among sensor- based data. More accurate results of the predictions of different metallurgical parameters with the exploitation of neural networks are based on the fact that the application of neural networks enables the assignment of relations among process parameters which cannot be traced using common methods due to their mutual interactions, the considerable amount of data, dynamics and the thus ensuing time demands. INTRODUCTION The thermovision technique has often been used in technical practice as a physical process for the optimization of some technological processes. Thermovision outputs are thermograms that are in the form of a quasi-coloured imaging record of a temperature field on an observed surface. The temperature of each detector element is transformed and digitalized. Correct temperature detection from detector acquired data is entirely dependent on many factors, for instance emissivity [1]. The paper [2] introduces a new method to improve the reliability and confidence level of defect depth measurement based on pulsed thermographic inspection by addressing the over-fitting problem. In [3], the authors measure heat fluxes on the rotating components of gas turbines. In this work, quantitative image processing tools were developed to perform fast and accurate infrared thermography measurements on 3D-shaped film- heaters directly deposited on the turbine endwalls. The newly developed image processing method and instrumentation were used to measure the heat load on the rotor endwalls of an axial turbine. The fatigue behavior of the AZ31B magnesium alloy subjected to three-point-bending tests was studied in [4]. The real-time recording of the surface-temperature-evolution data of specimens was obtained using the infrared thermography technology during fatigue. The characteristics of heat production and transfer were described and the correlation between temperature and stress distribution was analyzed. A key aspect in assessing the thermal standard of building envelopes is the quantification of the heat loss though thermal bridging, which can be expressed in terms of the linear thermal transmittance [5] This study shows how the infrared thermography technique can be used as a non-invasive and easy-to-use method to provide quantitative measures of the actual thermal bridging performance. The novelty of this approach includes the evaluation of the actual heat flow rate caused by the thermal bridge. Crassulacean acid metabolism is an alternative carbon fixation pathway that imparts high water-use efficiency in plants adapted to warm, semiarid climates. To facilitate this research, new tools which simplify procedures for detecting and measuring CAM are needed [6] An original approach based on active Infrared Thermography addresses the very challenging issue of the non- destructive analysis of bonded, repaired laminates. [7] Difficulties come from the weak property contrast between parts of repaired assemblies and the thickness of the joint. Strong attention is given here to the control of experimental test conditions (heat load, boundary conditions), which enable developing a physically consistent numerical model of the thermal problem. The most recent literary and original experimental results collected from different research and practical areas are presented in [8]. It is shown that the modern infrared technique acts as an indispensable laser partner in various biomedical and many other applications and technologies. And vice versa, the laser-based methods and techniques often serve as an appropriate research instrument enriching infrared technique measurement data with independently obtained information. In our work, we present ANN as a tool for the analysis of thermograms. The presented method also enables carrying out 13th International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics 408