Automatic Focusing Algorithm For Sea Aerial Photography Based On Saliency Detection

Aerial imaging has been widely used in various fields due to its large ground imaging area and flexible imaging time.Reflective waves,fluctuations in the height of the seabed and other factors will cause greater interference in the auto-focusing process,thereby affecting the accuracy of the auto-focusing.Therefore,the saliency detection can position the focusing window with the high-value target area accurately,which can improve the imaging quality of the target area.Focusing on the selection of the focusing window,this paper puts forward the problem of poor robustness of the basic algorithm of saliency detection,and proposes two improved saliency detection algorithms.Aiming at the situation of multiple targets in one scene,a value judgment method based on the shape and size of artificial targets is proposed.Aiming at the relative movement in aerial photography,a registration algorithm for the focus window is introduced.In the research of focus evaluation function,this paper analyzes the index of the focus evaluation function firstly.Aiming at the impact of illumination changes on focus evaluation,a focus evaluation function based on gradient matrix with correction of illumination is proposed.Aiming at the problem that the frequency-domain focus evaluation function cannot take the effective range and accuracy into account at same time,a frequency-domain focus evaluation function with a variable effective range is proposed.In order to monitor the in-focus state of the camera during shooting,a sharpness evaluation function without reference image is used.The experimental results show that the algorithms and strategies proposed can effectively solve various problems in the process of automatic focusing of aerial imaging.