The Airport Background Model Based On Gaussians Mixture

Airport surface surveillance is an important task to assist the civil aviation operation. Because there are some shortcomings of other methods, we decide to use the binocular stereo vision equipment to complete the airport surface surveillance task. The airport surface surveillance task is divided by 4 parts which are the detection, identification, tracking and behavior understanding of the target. The target detection, which is the first step, is an important and basic task. The objective of target detection is to split off the changed region which is generated by the target movement from the background. That is the reason why it's meaningful to split the target and the back ground. First building up the model of the background and then cutting it off is a useful method to do the target detection, so it is significant to build a proper model of the background.This article is about how to build up the model of the airport background and then to detect the target. The main contents of this article are as follows:1. Study various background modeling methods then choose the best method to do the airport back ground modeling, with the reasons of the choice.2. Build up the Gaussian Mixture background model with the knowledge of the statistics. Test its performance and find out its shortcomings.3. In order to make up the defect of the large amount of computation, we modify the Gaussian Mixture background model. The modified method can enhance the efficiency of the calculation.4. In order to solve the problem of the low detection accuracy, modify the algorithm and use the mathematical morphology ways to improve the results of the detection.