Методика визначення розливів нафти в Керченській протоці за різночасовими космічними знімками

Abstract

The aim of the work is to familiarize with the method of satellite monitoring of oil spills on the sea surface, the formation of skills and abilities to detect the presence of changes on time-varying space images for a specific area with the help of DeltaCue module ErdasImagine software package. Method. The technological scheme proposed in the paper involves the use of two independent methods of processing time-varying space images to determine the oil spill area in the Kerch Strait and compare the results. The first technique is to use of defining changes to DeltaCue objects, included in the ErdasImagine software package. This module includes the processing of images before and after the event. When processing data, it is possible to use three filters: spectral segmentation, incorrect recording of pixel pairs images and spatial filtering. The second method used is the uncontrolled classification by the ISODATA algorithm. Firstly, an image of the oil spill was classified, and then after the spill. With an image after the spill it was carried out 6 iterations, while not grouping classes to two: spill and everything else. After that, the results were exported to a vector format that detected the difference in the images at the spill location. The obtained data is comparatively and it is determined that the method using the DeltaCue module showed a better result and is more suitable for this kind of work. Results. As a result of the research, the oil spill area in the Kerch Strait in 2007 was received. Using the DeltaCue module of the ErdasImagine software package, an area of 30,918.4 km2 was obtained. After taking an uncontrolled classification of images, an area of oil spill was obtained at 29717.5 km2. After analyzing the results obtained, it can be concluded that for such studies, it is more efficient to use the DeltaCue module, since in the uncontrolled classification a part of the pixels is recorded incorrectly. Scientific novelty. The methods presented in the studies are well-known, and the scientific novelty lies precisely in their comparison on the example of a man-made disaster. Practical significance. Using these techniques, you can determine the effects of various natural and man-made disasters and determine which one is best suited to one particular case.