Section: Methodology and research methods. Models and forecasts
We use the remote sensing data demanding appropriate processing and special mapping technolo-gies. A mathematical model of image processing is proposed, based on numerical optimization of a multi-parameter nonlinear function with linear constraints with respect to the standard, the latter be-ing set by the expert for the selected image at the beginning of processing. As the image, the satel-lite imagery data, obtained by the LandSat-5 satellite during 10 years since 1987, are used. The pa-rameters of the measuring function of the technical vision system, which also performs numerical measurements of the water body, are determined. Due to the multimodality of the objective func-tion, there can be variety of solutions. The expert keeps freedom in selecting a set of parameter val-ues. To determine a set of parameters for the measuring function of the technical vision system, we carry out "calculation of the lake area" and data collection. The data are processed by algebraic and statistical methods. The results of the study showed that the lake area increased by 2.5-3 qu.km. We suggest that one of the reasons of changing the lake area is tectonic activity.
Keywords: lake area, satellite image processing, Landsat 5, vision system, remote area measurement, modified descriptive image algebras.
Article published in number 1 for 2021 DOI: 10.25750/1995-4301-2021-1-058-064