Human computer interactive image processing applications

The aim of study is related to giving commands and influence of computer with hand gestures which is the foreground object in real-time image in flux. The aim of the thesis is to give commands to computer with hand marks from a video image in real-time. Background extraction comes to the fore at the top of the most important methods used to determine the moving foreground objects in real time image. When simply considered, background extraction accepts and then exterminates the same pixels in the sequential image captures as the background, detects just different pixels and is based on the detection of the moving object. However, with this method, as even a tree in the wind shaking will be determined as the act, algorithms updating the background information continuously have been developed. In this study, an open source image processing library OpenCV and background extraction methods were examined. After the methods have been examined and due to the efficient use of memory, the production of fast results, its working in ambient conditions, and its application in real-time video displays, the availability of background extraction methods Code Schedule (Codebook) method was applied.