A method and system are disclosed for tracking objects which are crossing behind a host vehicle. Target data from a vision system and two radar sensors are provided to an object detection fusion system. Salient points on the target object are identified and tracked using the vision system data. The salient vision points are associated with corresponding radar points, where the radar points provide Doppler radial velocity data. A fusion calculation is performed on the salient vision points and the radar points, yielding an accurate estimate of the velocity of the target object, including its lateral component which is difficult to obtain using radar points only or traditional vision system methods. The position and velocity of the target object are used to trigger warnings or automatic braking in a Rear Cross Traffic Avoidance (RCTA) system.