The Research On Auto-focus Algorithm Of Space Camera Based On Image Processing

Autofocus technology has a wide range of applications in optical imaging systems.The performance of the focus system determines whether the imaging system can capture clear images.For the space camera,the traditional focusing method has various limitations,and the automatic focusing technology based on image processing has the advantages of simple principle,fast focusing speed,high focusing accuracy,wide application range and high integration efficiency.It is receiving widespread attention and research.By analyzing the characteristics of the space camera imaging of moving targets,it can be found that,unlike ordinary digital cameras,the target distance is far,the imaging target is smaller in the field of view,the blur of the out-of-focus image is larger,and because of the observation target and the observation platform.There is a relative motion between them,and there is an image shift in the target image point.Therefore,in order to solve the impact of these problems on the auto-focusing system,the following content was studied.(1)The imaging characteristics of the space camera are analyzed.The space camera has the following problems in the imaging process: the observation distance is long,the effective pixels are small,the contrast of light will cause the yin and yang surface to be generated,and various noises will be introduced in the case of low illumination,and the moving target Motion blur will occur.Therefore,it is necessary to properly preprocess the image and analyze the effect of motion image shift on focusing.(2)The image sharpness evaluation algorithm was studied.The focus evaluation function selection is the key to the auto-focusing technology.The purpose is to separate the focus image and the focused image,determine whether the image is out of focus,and then find the position where the image with the sharpest image is located,and complete the auto-focusing operation.This paper presents an improved Brenner function with varying coverage.This function uses a simple Brenner function as a calculation template,and uses pixel blocks instead of a single pixel to perform gradient calculation.This method can effectively solve the problem that the focus range is small due to the large amount of defocus blur,and because it is The operation of the pixel block can effectively eliminate the influence of isolated noise points.The experimental results show that the proposed algorithm effectively improves the noise immunity,and the sensitivity of the focus adjustment is greatly increased in the area where the defocus blur is too large,which effectively improves the focus search area.(3)The method of constructing the focusing window was studied.In order to quickly and accurately select the region where the observation target is located,this paper uses the saliency region extraction algorithm based on the minimum obstacle distance transform to extract the saliency region of the image and construct a rectangular focusing window with adaptive target changes.Through simulation experiments,it can be found that the focusing window constructed by the method proposed in this paper can effectively reduce the computational complexity of the system,eliminate the influence of background noise,and improve the accuracy and real-time performance of the focusing system.(4)The focus search algorithm was studied.This paper introduces the characteristics of the traditional focus search algorithm and its shortcomings in practical applications,and proposes an improved mountain climbing search algorithm.This algorithm first adjusts the focus in large steps,and then fine adjustment in small steps can effectively improve the search speed and accuracy,and reduce the impact of local extremes.