Labeling of names in the video stream

• Main Author: Li, Hui Ting.
• Other Authors: Teoh Eam Khwang
• Format: Final Year Project
• Language: English
• Published: 2013
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/52604
• Institution: Nanyang Technological University

It is very desirable for media users to know the names of people in videos. Naming persons who are appearing in video with face detection and extracting distinctive invariant features can help to research and manage the video content. In this project, two algorithms will be developed to label the names of the people appear in the video stream by using k-nearest neighbor algorithm and support vector machine (SVM) algorithm. Firstly, the training data sets of Jackie Chan and Non-Jackie Chan must be prepared. The video needs to segment to frames and pick up those key frames as the test images. Secondly, all the training images and test images go through with face detection program and feature extraction program. Lastly, the results of the two classification approaches by using the KNN and SVM were discussed and compared. Recommendations will be mentioned in this report as well.