A rough fuzzy neural network approach for robust face detection and tracking - Emerging Trends in Image Processing, Computer Vision, and Pattern Recognition - page 249

and w k and v k are the zero-mean, white, Gaussian random processes modeling the system
noise. The whole filtering procedure consists of prediction and correction steps, which are car-
ried out alternatively. The state in the next time step is predicted from the current state using:
Face detection is performed on the window of height H k + 1| k and width 0.75 H k + 1| k centered at
a predicted position. Detection within the window instead of the whole image helps to reduce
the detection time, which is important for real-time operations. The Kalman corrector is
where K k + 1 is the Kalman gain, computed as
The covariance matrix P is updated as follows: