基于运动特征和中心环绕模型的空对地车辆检测方法

平台的移动基于空基车辆检测以其视野广、速度快、适应性强等优点,近些年成为了智能交通和航空应急救援等领域内的一个热点课题。然而,在城市背景的交通环境下,由于复杂背景造成的干扰、车辆运动的不规律性以及光照等条件变化带来的噪声,有效地获取运动对象特征是空对地视觉监视方法面临的一个难题。为有效解决这一问题,提出了结合运动特征的中心环绕模型进行空对地车辆检测的算法。现有的空对地车辆检测算法,大多是针对固定的摄像头的情况下,基于运动特征进行检测,例如光流法、背景差法等。这些方法在对高速公路等背景相对较为简单的环境下获得了良好的检测结果,但当面对城市复杂背景条件或摄像头自身运动的状况时,难以有效地检测出运动目标。而在对颜色、强度等静态特征进行显著度分析的基础上,加入了对运动特征的分析,可以解决在复杂环境下准确提取车辆目标有效特征的问题。试验表明,相较于基于图像相减的检测算法和单一显著度算法,该算法可以获得更高的检测率与更低的误报率。

The Method of Airborne Vehicle Detection Based on Motion Feature and Center-surround Model

In recent years,airborne vehicle detection system has been a hot research topic in many fields including intelligent transportation system and airborne emergency rescue because of its advantages such as wide view field,high speed and good adaptability.However,in the context of urban traffic environment,due to the interference caused by a complex background,vehicle movement irregularities and noise caused by changes of light,airborne detection system is facing a problem of the method to obtain the visual characteristics of moving objects effectively.In this paper,center-surround model combined with the motion feature is proposed to solve airborne vehicle detection.Existing detection algorithms,which are mostly for stationary camera systems,usually detect based on motion features,such as optical flow method,background subtraction method.These methods has gained and good results in relatively simple traffic background such as the highway,but when faced with complex urban background conditions or the status of camera platform movement,it is difficult to effectively detect moving targets.In this paper,analysis of the motion feature is added to the static features abstracted from the color and intensity characteristics,so that the problem of the effective and accurate extraction of the target features in a complex environment can be solved.Experiments show that,compared to image subtraction based detection algorithm and the single saliency algorithm,this method can achieve higher detection rate and lower false alarm rate.