Road sign tracking for adaptive cruise control under nonlinear conditions

A vision-based system is described for adaptive cruise control to track road signs from within a moving vehicle. The proposed system has the standard architecture with a particle filter due to its robust tracking performance in a complex environment. In the case of tracking road signs in a real environment, it has great difficulty in predicting time series data by reason of its nonlinear characteristics. To overcome this problem, proposed is a state transition model to which the differences of affine parameters and a fuzzy-autoregressive model are applied. Also, to make an observation model, the Parzen window is used at a measurement step in the particle filter.