NATURAL SCIENCES

Journal of fundamental
and
applied researches

Face detection

2013. №4, pp. 199-215

Mohammed Mamdouh Mohammed Gomaa - Assistant Lecturer, Minia University, Minia, 61519, Arab Republic of Egypt, doha_gomaa@yahoo.com

In the development and analysis of image processing systems useful and often necessary to have a mathematical description of a subject image processing. The task of face detection in the image (face detection) is often the “first step” in the pretreatment process of solving the problem of a “higher level” (e.g. recognition face recognition facial expressions). However, the very presence information, and possibly the number of people in the image or the video stream can be useful for applications such as security systems and meaningful indexing databases of images or video clips. In the context of building a natural human-computer interface problem localization person finds its application in face tracking subsystem and its characteristic features in the video for the recognition of facial expressions, to determine its position in 3D space to generate computer commands. In machine vision often two versions of the face detection problem – Localization person (face localization) and tracking of faces (face tracking). Localization person is a simplified version of the problem of detection, as based on the knowledge that the image there is one and only one person. Task of tracking the movement of the person in the video stream can be formulated as a problem of localization of persons at the current frame based on information about its position on the previous frame. In this paper we present survey of face detection algorithms.

Key words: Face Detection,Face Localization,Skin Detection,Color Space,Neural Networks Viola-Jones Method,