Abstract: A computer vision solution scheme for the automatic landing of an unmanned helicopter on ship was examined from the view of engineering application. A series of computer vision algorithms such as Sobel filter, FastHough Transform, were adopted and integrated to get the relative positions between the helicopter and ship, fromwhich their relative movement parameters can be obtained. The image segmentation was executed in HSV colorspace. Hough Transform was adopted to detect the linear parameters, the bottle neck of which was broken with theuse of gradient information. The use of Hough Transform makes the stereo vision scheme very robust to noise.Computer simulation results indicated that the presented algorithms were robust and reliable.

Key words: unmanned helicopter, computer vision, image segmentation, edge detection, Hough transform