Face Detection System / Chan Beh Teng

Chan, Beh Teng (2004) Face Detection System / Chan Beh Teng. Undergraduates thesis, University of Malaya.

Preview

PDF (Academic Exercise (Bachelor’s Degree) Download (33Mb) | Preview

Abstract

The aim of this project is to replicate on a computer what humans beings are able to do so effortlessly - detect the presence of the faces in their field of vision. To a layman, it may appear trivial, but to actually implement the necessary steps leading to successful execution of this in n algorithm is difficult and still an unsolved problem in computer vision.The proposed Face Detection System is an application that automates the detection of frontal human faces(s) in a given image. The System perform all tasks including image pre-processing, analysis and output. The user merely present it with the image. The system takes as input a 24 bit RGB colour image which containts face(s). Before the detection process begins, the system pre-processes the image into a format required by the system. The system will generate an output where every detected face will be indicated with a rectangle on a copy of the original image. The method used for detecting face is a combination of color based technique and template matching. A gray-scale image is generated from the original input image. Skin regions are then separated from skin regions by segmenting the gray-scale image. After that, the system will locate the frontal human face(s) within the skin regions, using template matching. The Face Detection System is implemented in visual Basic 6.0 The system was developed based on the Unified Process,which uses object-oriented paradigm as the basis of software development.

Actions (For repository staff only : Login required)