Study And Design On FOD Detection And Recognition Algorithm And Surveillance System For Airport Runway

With the fast development of aviation technology and civil air transportation industry, the number of civil airport and passenger greatly increased in recent years. Foreign Object Debris (FOD) would potentially cause huge damage when it appears on the airport runway. Due to the airport's complex circumstance, FOD detection and surveillance system could detect, analyze and give the warning to the FOD target on the runway quickly and precisely. The FOD surveillance system is one of the important protections for airplane's safety now.The multi-sensor target detection and recognition technology is integrated with the technology of multi-sensor information fusion and traditional target recognition. Multi-senor target detection and recognition is a paradigm of information fusion technique dealing with target detection problem, it combines multi-sensor data and extracts colorful target features to earn a better detection and recognition performance than the single sensor does.Normally, the FOD detection and surveillance system contains millimeter wave radar, visible and infrared light sensors, which have their own advantages. The millimeter wave radar can measure accurate target angle and range but its signal contains less data. The visible light sensor can provide more image details and enrich the description of the target feature but its performance is vulnerable to the background and weather condition. The infrared light sensors have great advantages in the night but its measure range is limited by the high cost.This thesis aims to study on the specialty of the multi-sensors in the FOD detection and surveillance system and their working assignment and data interaction. This paper's concentration is studying the FOD detection and recognition algorithm based on visible and infrared images. At first the foreign FOD detection and surveillance system and technology have been discussed and analyzed, then a new FOD detection and surveillance system including millimeter wave radar, visible and infrared cameras has been designed.After that, according to the characteristics of images of airport runway a quick and effective detection algorithm based on the Canny operator and background model is introduced. Then the content-based image features including gray, texture and shape are extracted and alternated to a designed threatened level. The classifier is constructed by the artificial neural network to estimate the threatened level. The simulation experiments indicated that the proposed method and algorithm was effective and efficient, satisfied the robustness and real-time response.At last, the FOD detection and recognition algorithms in this paper are applied to the"FOD detection and surveillance system simulation software", and the experiment results are according with the system requirements.