Registration Number
0218U006222,
0112U008453
, R & D reports
Title
Contours Extraction for Structural and Statistical Object Identification
popup.stage_title
Head
Krylov Victor Mykolayovych, Доктор технічних наук
Registration Date
03-04-2018
Organization
Ternopol National Economic University
popup.description2
. The research object is the process of visual information processing and decision-making under uncertainty conditions in automated video surveillance systems. The aim of the work is to increase the speed of the contour extraction methods and to characteristic point's selection for identifying individuals in applied intelligent information technologies for computer recognition in limited computing resources. Methods of analysis are: methods of system analysis, methods of contour extraction, wavelet analysis, functional analysis, methods of mathematical statistics, methods of modeling and programming. The following research results were obtained: analysis of the existing video surveillance systems, methods of contour extraction and methods of contour characteristic points selection; a method and basic information technology for contour characteristic points selection with object adjustable detailing, based on the wavelet analysis of the curvature function, which reduce the number of contour points in accordance with the processing purpose; the method and basic information technology were improved; they are characterized by the use of wavelet functions to highlight image intensity changes that allowed to obtain a sequence of the image object hierarchical shapes with the adjustable detailing. The obtained results can be used in video surveillance systems to improve its performance
Product Description
popup.authors
Антонюк (Брушніцька) Анастасія Сергіївна
Биковий Павло Євгенович
Загородня Діана Іванівна
Ковалок Костянтин Юрійович
Комар Мирослав Пертович
Палій Ігор Орестович
popup.nrat_date
2020-04-02
