一种基于类间方差地平线检测算法.pdfVIP

31 10 Vol31 No10 2010 10 ACTA AERONAUTICA ET ASTRONAUT ICA SINICA Oct. 2010 :2010) , , , , ( , 100081) A Horizon Detection Algorithm Based on Betweenclass Variance Analysis Cheng Xu, Hao Qun, Song Yong, Hu Yao, Zhang Kai ( School of Optoelect ronics, Beijing Institute of Technology , Beijing 100081, China) : ( MAV) , , , , : , 1, 3 , 198% 495% , : ; ; ; ; : V24932 : A Abstract: To detect the horizon for the steady flight control of micro air vehicles ( M AV) in vision navigation, a horizon detect ion algorithm based on betw eenclass variance analysis is proposed. T he angle and distance are used as parameters t o exhaust the st raight lines in the image. T he blue color component of the pixel is chosen as the feature for dist inguishing the sky and the ground. A criterion based on the betw eenclass variance is set up for determining w het her or not a straight line is the horizon within the exhaustive straight lines. Experi mental results show that, compared w ith the horizon detection algorithm based on color covariance mat rix cal culation, the proposed algorithm performs better . T he average detect ion error decreases by 1 degree in angle and 3 pixels in distance. The correct detection rate increases by 1. 98% in angle and 4. 95% in distance at a sat isfactory speed. Key words: m icro air vehicles; comput er vision; image processing; attitude cont rol; horizon detection [ 4] ( Micro Air Vehicle, MAV ) , [ 5] , , ,