Research Of Quantum-MeanShift Algorithm Base On The Moving Multiple Vehicles Tracking Of Video Information

As an important part of intelligent transportation, vehicle tracking has been widely used in the field of transportation, such as traffic flow monitoring, vehicle navigation and auxiliary safe driving, etc. It greatly improves the vehicle safety, comfort and road utilization rate. With the continuous development of video information technology, video monitoring technology has been known as the main means of monitoring the traffic management lies in its intuitive, with features of visual information and wide coverage. Video monitoring technology has a wide range of applications. Since the randomness of the movement of the vehicle and the complexity of the actual scene, it is difficult for vehicle tracking in real-time, accuracy and stability of the system. So this paper presents quantum-meanshift algorithm. The main work is as follows:1. Proposed moving multiples vehicles tracking algorithm based on Mean-Shift. It builds a mathematical model for mean shift algorithm based on moving multiples vehicles tracking. Through the simulation analysis of moving multiple vehicles tracking based on mean shift, the limitations of the mean shift algorithm itself is obtained.2. By introducing the quantum evolutionary strategies, we proposed moving multiple vehicles tracking algorithm based on quantum-meanshift. Using the quantum evolutionary algorithm to search the optimal location of vehicle in the current frame, and calculate the mean shift algorithm similarity. The simulation result has proved that this algorithm has better accuracy and stability than mean shift algorithm, thus, this algorithm is effective.3. By introducing the Kalman filter, we proposed an anti-occlusion algorithm for moving multiple vehicles tracking based on quantum-meanshift. It can determine whether a serious occlusions using Kalman filter. When a serious occlusion turns out, we use Kalman search strategy to confirm the position of the vehicle. The simulation result has proved that this algorithm can solve occlusions problem well. Thus, this algorithm is effective.4. It analysis of the multiple cameras relay in the process of multiple cameras synchronization tracking and target handover problem, and introduce the target handover algorithm based on FOV line. Combined with the quantum-meanshift algorithm, we proposed multiple cameras relay algorithm for moving multiple vehicles tracking. Using the FOV lines to determine whether vehicle is in public area. When the vehicle is in the public area, it identifies the target by target corresponding relation. The simulation result has shows that this algorithm has well accuracy, thus, this algorithm is effective.