Traffic wave real-time detection and tracking method based on information fusion

The invention relates to a traffic wave real-time detection and tracking method based on information fusion. Firstly, a traffic video image is acquired, then a Kalman filtering technology is adopted, and the traffic wave state prediction value at the current moment is acquired according to the traffic wave state optimal estimated value at the last moment; then, the traffic wave position measured value at the current moment is obtained, and a double adjustable window algorithm fused with an Adaboost classifier is adopted in each vidicon in the visual area between the opposite vidicons; estimation is carried out based on a feature point tracking method and a traffic wave estimation method in the blind area of the vidicons away from each other; finally, a weighted average consistency information fusion algorithm is adopted to obtain the traffic wave state optimal estimated valve at the current moment, preparation is made for the traffic wave state prediction of the next moment, and in the process, the measurement noise weights of the vidicons are introduced, and the detection result reliability of each vidicon is dynamically adjusted to improve traffic wave locating accuracy.