Object detection method for a vehicle driver assistance system based on maximum a-posteriori estimation of Bayesian probability functions for independently obtained measurement values

摘要

The method in which a video image sequence is evaluated by maximum a-posteriori (MAP) estimation of Bayesian probability functions to provide an event value which describes the state of an object (20). The method involves the following steps: image evaluation to obtain at least two independent measurement values for the event, assignment of a quality factor to each measurement value; modeling of the probability distribution of each measurement value using a gauss distribution with the quality factor and formation of the product of the gauss distribution as the Bayesian probability function. An independent claim is made for a device for object determination in a driver assistance system with a video camera and an object determination module.