基于混合高斯模型的跑道侵入检测方法

Runway Incursion Detection Method Based on Gaussian Mixture Model

摘要:

面对机场不断增加的航班量,为了保障航班安全,避免发生跑道侵入,使用了基于混合高斯模型的背景减除法,在机场采集的视频数据中,追踪跑道安全区附近的运动目标。使用Graham Scan算法计算动态目标像素点的边框,根据前后帧之间动态目标的变化,计算其运动状态。在视频图像中标记跑道及跑道安全区,并使用旋转卡壳凸包算法计算动态目标与跑道安全区的最小距离。提出了的跑道侵入检测模型及方法,用于检测是否存在跑道侵入风险。仿真结果表明,该方法能够有效的机场安全区附近的追踪动态目标,并评估跑道侵入的告警等级。

Abstract:

In order to ensure flights safety and prevent runway incursion, Gaussian mixture model(GMM) based video background subtraction is used to track moving targets near runway safety area from video collected at the airport. The convex hull of the moving target’s pixel point set is calculated by Graham Scan algorithm and the state of motion of the moving targets is calculated by different between video frames. The distance between the moving targets and runway safety area is calculated by marking runway and runway safety area and Rotating Calipers algorithm. The runway incursion detection model and method are established to evaluate the safety status of the runway area. The simulation results show that this method can track moving targets near runway safety area and evaluate the alarm level of runway incursion efficiently and effectively.