Automatic extraction and tracking of face sequences in MPEG video

Abstract

This PhD work focuses on the problem of extracting multiple face sequences from MPEG video based on face detection and tracking. It aims to facilitate the strata-based digital video modelling to achieve efficient video retrieval and browsing. The research includes the design of efficient algorithms for face detection and tracking, and tools for compressed domain video processing. It differs from existing efforts in that it focuses on developing efficient techniques to make use of the features in DCT domain and the characteristics pertaining to the image and video compression standards, such as JPEG, MPEGs, H.26x, etc. In summary, the research work consists of: 1. A system to extract multiple face sequences from MPEG video. 2. A DCT-domain approach to face detection in MPEG video. 3. DCT-domain approach to fractional scaling and inverse motion compensation on image and video.