High Speed Face Detection And Tracking
Real time face detection and tracking is one of the most important components used in many image processing
applications. Gesture analysis of a driver driving a vehicle based on facial features like eyes, and mouth is one such
application we are considering. The image frame rate needed in such applications is at least 15 frames/sec say for example
on a smart phone where our algorithm resides on application layer. In this paper we focus on Joint Viola Jones face detection
and tracking algorithm which considers some innovative techniques to speed up processing time. It should be noted that
entire process of gesture analysis is automated with high degree of robustness with a view to alert the driver at appropriate
time to avoid mishaps. The performance analysis of Joint Viola Jones face detection and tracking algorithm is analysed using
functions available in Computer Vision Tool box and Image Processing tool box of MathWorks. The performance is
platform dependent. In the end we show performance results, which is quite promising.
Keywords- face detection, tracking, Mean Shift algorithm, Viola-Jones algorithm.