A Correlation Based Approach for Velocity Estimation in Real-Time Person Scanning

We want to present a method to estimate the velocity of a moving person in security scanning based on a correlation approach for subimages. The velocity estimate will give a signifikant gain in image quality and is performed without any additional sensors using only the imaging system. It is shown and verified that a high precision velocity estimate may be achieved by thinning out the full image to 1/25 of its original size.