Automatic Runway Recognition And Tracking In The Sequence Airfield Infrared Image Field

This thesis is emphasized on the research of automatic recognition and tracking technology in the sequence airfield infrared image field. As an important part of infrared tracking, a great many scholars have exerted themselves. An improved algorithm of runway recognition and tracking in the sequence airfield infrared image field is proposed on the basis of summarizing and digesting the existing concerned research result.Starting with image pretreatment, infrared image denoise methods had been researched. Then airfield image segmentation was discussed mainly on the threshold segmentation and edge detection. Combined with the airfield fabric characteristics, a mathematical model was established which paves a way for the runway recognition and orientation.The beeline extraction technology has been mainly studied. The chain code tracking, classical Hough transformation, random Hough transformation, RHT-GD and Radon transformation were then discussed. Ultimately, a novel Hough transformation based on chain code tracking was carried out. Simulation results proved its superiority.The underlying purpose of the research work was aimed at orientation and tracking of runway. An airfield fabric model had been introduced in, by which the runway had been located accurately. Additionally, a new algorithm that confirm the tracking flow utilizing area of runway in the infrared image was applied to ensure the stabilization, precision and speediness of tracing course.