Satellite imagery obtained from open sources intelligence is of great importance for the security of the state, as the data obtained allow monitoring the location and activity of troops, as well as detecting and identifying objects and military equipment. The evaluation of the quality of these data is an important factor in their practical use. In the world literature, many authors considered the problem of image quality in terms of the spatial resolution assessment, but in this article the research concerned determination of the imagery interpretability and its improvement. The article presents the methods of evaluations and qualitative analysis of reconnaissance imagery. The analysis were carried out both by subjective and objective methods, in order to optimize the individual approach of specialists to the problem being addressed. The subjective assessment method consisted in the visual analysis of the views, which were detailed in the NIIRS. The objective evaluation was made based on the calculation of the metrics characterizing the spectral and spatial quality of the imagery. In addition, the influence of various signal processing methods was studied to improve the quality and potential of image interpretation. Radiometric amplification operations, context transformations and pansharpening were applied. The conducted research work have allowed the development of the concept of methods for improving the image quality, which resulted in their better interpretation. The use of signal processing methods (mainly radiometric amplification and high-pass filtration) resulted in improved image quality in the assessment based on both mathematical indicators and the NIIRS.
The subject of study were low resolution SAR imagery, provided by Sentinel 1A and 1B, which can be obtained using open sources. Due to the great interest of products, an analysis of the interpretation possibilities of imaging acquired in interferometric wide swath mode and extra wide swath mode were made. Before exploring the possibilities, the proposed methods have been tested to improve the quality of radar images. Depending on the analyst’s needs as well as the polarization channel used, adaptive filters such as Frost, Lee-Sigma, and in some cases Gamma-Map are recommended. However, the use of classification and pseudocolor allows dividing the area of interest into basic fields, such as the urbanized area, water or vegetation. Satisfactory results were also obtained by integrating different polarization bands. The impact of incidence angle of radar beam on photographing the object was also shown.