Reference image quality assessment of composite materials exposed to positive and negative temperatures

ABSTRACT Introduction. The question of a comparative assessment of the quality of images obtained as a result of scanning test composite materials exposed under the cyclic effects of positive and negative temperatures is considered. The samples are sorted relative to the reference numerical metric expressed as a percentage. Materials and methods. The compared composites differ in composition relative to the control sample. Periodic scanning of the tested composites is carried out during several control points of time. A comparative analysis of the image quality of test composites was carried out both during the entire exposure time and at the control points of time. On the basis of the existing algorithm Structural Similarity Index for measuring image quality (SSIM), percentage indicators of sample quality were determined depending on the composition and exposure time. The results obtained allow the selection of composite materials in accordance with the level of structural similarity in structural color characteristics. According to the SSIM algorithm, the resulting metric values are in the range [-1; +1]. Full structural similarity corresponds to a value equal to one. It is proposed to express the SSIM metric as a percentage with the transformation that the change in the color structural difference of the relative control sample can vary from zero to 100 %. Results. For the proposed method of comparing images of scanned samples of composite materials, ten samples of composite materials of different composition were used. Numerical results of the experiment with their graphical visualization give a clear idea of the nature of the change in color characteristics relative to the control sample. Consistent evaluation at the exposure points shows the nature of changes in the properties of materials from the duration of exposure to cyclic changes in positive and negative temperatures. Conclusions. The proposed benchmark image quality assessment based on the existing SSIM algorithm allows you to quickly investigate groups of materials, the change in properties of which is caused by their exposure in adverse operating conditions.