Analýza typů částí oděvu osoby z obrazového signálu

Abstract

Bachelor thesis deals with matter of image processing and analysis of the types of clothing details from image signal. Fundamental goal of the thesis is to create algorithm capable to detect protective propts on single person. The paper consists of two parts. First part enlightens digital representation of image and its processing, describes detection methods and identification of particular objects. Second part comprises the progress issue itself and parses the algorithms designed in MATLAB development environment for detection and identification of protective helmet based on color segmentation and size ratio of the propt. Devised algorithm is described within steps with displayed partial results. Visualization of designed algorithm in graphical user interface GUI is also inhered to the thesis.