Search academic texts

Information × Registration Number 0221U106729, 0121U113956 , R & D reports Title Machine learning based processing of multichannel remote sensing data popup.stage_title Head Uss Mykhailo L., к.т.н. Registration Date 18-12-2021 Organization National Aerospace University "Kharkiv Aviation Institute" popup.description2 Scientific research report: 72 pages, 27 figures, 11 tables, 123 sources. The object of research is multi-channel images that are formed and processed by remote sensing systems. The purpose of research is to develop machine-based methods for efficient automatic and semi-automatic processing of multichannel remote sensing data. The report presents the results of the development of methods for finding similar blocks (fragments) for component images of multichannel remote sensing data using combined similarity metrics optimized on the basis of machine learning. The analysis was performed for component images of the visible, infrared and radio bands, as well as digital terrain maps. Combination has been shown to significantly improve the reliability of search and subsequent matching. Also developed a method and tool for predicting the effectiveness of filtering for radar images distorted by spatial-coded speckles. Forecasting is based on neural network training, the inputs of which are dozens of statistical parameters that are quickly calculated for each image. This allows a reasonable decision about the optimal parameters of the filter. MULTICHANNEL IMAGES, MEASURES OF SIMILARITY, COMBINATION, FILTRATION EFFICIENCY, CLASSIFICATION, NEURAL NETWORKS, LEARNING, TEACHING, TEDVED TEST Product Description popup.authors Abramov Sergіy K Vasilyeva Irina K. Lukin Volodymyr V. Proskura Galina A. Rubel Andriy S. Khutornenko Sergiy Volodymyrovych Chernova Halyna A. popup.nrat_date 2021-12-18 Close