An automatical labeling aircraft method for airport video monitoring
1.College of Civil Aviation, Nanjing University of Aeronautics and Astronautics, Nanjing, Jiangsu 210016, China; 2.Research and Deve-lopement Center, The Second Research Institute of CAAC, Chengdu, Sichuan 610041, China; 3.School of Electronic Engineering, University of Electronic Science and Technology of China, Chengdu, Sichuan 610054, China
Abstract：
Abstract: To solve the problem that the airport video monitoring could only provide target images without target labels, a novel method of automatical labeling aircrafts in airport video was proposed. The aircraft targets were detected and tracked in video image. Four and more correspondences were selected in video image and airport map respectively to compute homograpy matrix. The airport map coordinates were transformed to image coordinates with homography matrix. The aircrafts in airport video were automatically labeled by data fusion of image target track data and ADS-B data. The coordinate transformation error caused by measurement error of image point correspondences was computed. The relationship between coordinate transformation error and point correspondences geometric distribution was analyzed. The experiments were completed based on the measured data of Chengdu Shuangliu International Airport. The results show that the method can realize automatical labeling aircrafts for airport video monitoring.
2013, Vol. 34 
