Research On The Method Of Auto Focus System For Remote Sensing Camera

Ground-based optical remote sensing imaging camera is one of the main application directions of satellites,with focusing technology as one of core technologies in the camera imaging system.As remote sensing camera is susceptible to defocusing due to such factors as temperature,atmospheric pressure and the change in the imaging distance in the outside environment,the focusing function is set in the course of imaging to ensure the quality of the imaging system.At present,the commonly used method is to obtain the defocus amount by calculation according to the ground imaging effect,and then manually set the focus function.Therefore,it is necessary to develop an autofocus technology to simplify the focusing process.So far,many domestic and foreign scholars have conducted in-depth research on the existing autofocus technology,but it is still a very difficult task to improve the accuracy and speed of focusing at the same time in the imaging process of remote sensing cameras.Digital image processing technology is a means to flexibly obtain the important information,this subject has conducted in-depth research on four types of image sharpness evaluation functions and focus search algorithms commonly used in digital image processing technology,and contrasts and analyzes their shortcomings in remote sensing camera applications through Matlab simulation.Sobel-Canny is proposed the two-focus method combining the two operators is used to make a second-focus judgment on the imaging target to ensure the accuracy of the focusing function of the remote sensing camera.In order to further improve the speed of the system’s focusing,improvements have been made to the problems existing in the hill-climbing search method.Finally,the experimental results of five focusing methods are compared and analyzed on the remote sensing camera focusing test platform.The experimental results show that and the speed of focusing has been improved one time,it can meet the needs of function when the remote sensing camera is imaging at a long distance.