Study On Small Obstacles Detecting System Of Airport Runway Based On Image-pattern Recognition

With the deepening of globalization, the current global economic development faster and faster, in order to meet the requirements of development, many demands about feed has been put forward on the transportation industry. Among them, the fastest growing air transport industry is a top priority, variations of the airport number increases year by year. Therefore guaranteeing the security of airport flight is one of the first tasks. The runway as the plane normal navigation platform, there are obstacles to secure the aircraft on the runway will be unpredictable threat. This thesis proposes a runway tiny obstacle image pattern recognition system, it can effectively detect obstacles on the runway and intuitive judgment the threat level. And it could remove the potential harm on the runway effectively.Firstly, this thesis introduces the background of the subject, knows the research status about the runway obstacle processing system at home and abroad. In this premise, this thesis introduces the multi-sensor image pattern recognition system, and explains its advantages. The threat level to intuitive is proposed. In order to follow-up function extension, the thesis designs a set of airport runway obstacle recognition system based on image pattern recognition. The system compared to using a single type of sensor track real-time obstacle detection system, high efficiency and flexible expansion system algorithm is proposed. The overall framework of the system from outside to inside is divided into information collection, information processing and information application blocks.In the information collection block, there is a variety of capture devices used. The radar scans the runway during 24 hours to determine the presence or absence of an obstacle, and then control the optical camera locking the target which is suspected the existing obstacles of target location, and fix camera. In order to make sure the system can work properly all day, the visible light cameras and infrared cameras all worked.In the information processing blocks, this thesis uses digital image processing algorithms. On image processing, denoising operator, edge detection operator, background model, background subtraction and other ways to get the information collection can be efficient and obtain useful information. The information from the location of the feature, geometric features, shape features and other information obtained through the neural network processing visual obstructions merger threat factor, and the information is transmitted to the application layer via LAN.In the information application blocks, a man-machine mutual interface is established by the Visual Studio 2010 platform MFC. Because of the object-oriented features of MFC, the interface can display the runway timely, the obstacle information database, obstacles of real-time updating and preservation, etc.This thesis proposes a slight obstacle runway image pattern recognition system to identify obstacles to improve the efficiency by the use of artificial subjective judgments according to the threat level of the division of the obstacles, reduce the unnecessary labor cost, improve the security of the artificial operation at the same time, reduce the time of the closure of navigation when conventional detection, improve the quality of the project inspection, reduced the potential threat.