Search academic texts

Information × Registration Number 0219U102274, 0119U100940 , R & D reports Title Intelligent system of complex automatic processing of visual multimedia data popup.stage_title Head Ieremeiev Oleg I., Кандидат технічних наук Registration Date 26-12-2019 Organization National Aerospace University "Kharkiv Aviation Institute" popup.description2 The subject of the study is visual multimedia data, including digital images and videos, which are processed, transmitted and used in modern information systems and networks. The object of the study are visual quality assessment methods of of multimedia data, assessment methods of characteristics of destabilizing factors, processing methods of multimedia data and methods of effectiveness predicting of these methods. The report proposes the concept of interoperability of different digital processing methods of visual data, the practical value of which is confirmed by further developments. The results of studies of complex noise characteristics and properties in real images are presented and a method of estimation of such noise characteristics on the basis of neural networks is proposed. It is shown that the availability of information of noise dispersion can further improve the performance of the proposed method for predicting image quality. Such visual quality assessments and metrics make it possible, in particular, to decide a feasibility of applying a particular stage of processing, to set the optimal parameters of the processing methods, and to predict the effectiveness of methods application prior to their implementation. This also applies to image filtering, which is a must-have for most real-world images. The method is developed and testing is performed for predicting the efficiency of filtering RSA images with speckle noise from SENTINEL-1 remote sensing system. Product Description popup.authors Abramova Victoria V. Zemliachenko Oleksandr M. Kozhemiakina Nadejda V. Naumenko Viktoriya V. Rubel Andrii S. Rubel Oleksii S. popup.nrat_date 2020-04-02 Close