UDC 528.854.4
ALGORITHM TO IMPROVE ACCURACY OF RADAR IMAGE OBJECTS CLASSIFICATION BY CRITERION OF BELONGING TO WATER SURFACE OR LAND USING DIGITAL ELEVATION MODEL
L. L. Kuznetsov, 2nd category engineer at the Research Institute «Photon», RSREU, Ryazan, Russia; orcid.org/0009-0005-3888-8599; e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
The article suggests an algorithm that allows to improve accuracy of radar image objects classification by criterion of belonging to water surface or land based on combining the results of primary classification with a class mask generated from digital elevation model (DEM) data. The main task is to create an algo rithm that generates a class mask by analyzing relative height of surface objects. This work is relevant due to the need to create and improve radar image objects classification algorithms to solve the tasks of Earth re mote sensing (ERS) systems. The results of the proposed algorithm utilization showed a high degree of elimi nation of false shadow mountain slopes classification into water surface class and are quantified at up to 21,7 square kilometers of correctly classified Earth surface area.
Key words: Earth remote sensing, radar images, digital elevation model, objects classification, water surface detection.