Real-time detection and tracking of light point

A method of detecting and tracking light points in realtime images was prop osed. Detection of light points was based on analyzing of the segmentation results and graylevel change in different directions, which can distinguish light points from reflector points. Tracking of light points is based on a daptive adjustment of gate size and gate position. In order to improve comp uting precision, the Cardinal spline functions are used as the interpolatin g functions. This method is feasible and the results agree with the relativ e experiment.