Research On Computational Imaging Algorithm For Large FOV High Resolution Optical System

Wide field of view(FOV)and high resolution is always a couple of imaging characteristics which are opposite and unity.Wide FOV offers overall information of the target field and high resolution contributes to the details.Imaging system with wide FOV and high resolution being able to acquire large observation field and resolving power is widely used in the area of Earth observation and in the micro aerial.The manufacturing level of the detector,the diffraction limit and the limit of optical design is the traditional optical system constraints.so the traditional imaging system is difficult to achieve the desired standard.Especially for objects and ensure uniform illumination in high resolution wide field situation is very difficult to achieve.This paper proposes a method based on block overlapping contrast enhancement algorithm and super resolution reconstruction computation imaging algorithm,in order to improve the brightness and resolution of the traditional optical system.The block overlapping contrast enhancement algorithm based on the imaging blocking and overlapped gray scale adjustment algorithm,this algorithm can changes the gray scale of the light and dark place,makes the image detail information can be reflected at the these area,and the image entropy of image(the details of the image information)and brightness uniformity can be increased by 10%～15%.and the advantages of this method can reduce the noise and distortion of color.The algorithm of super resolution reconstruction based on sparse representation based on the imaging blocking and sparse representation dictionary learning theory,firstly discusses the method of dictionary learning and optimization method,then discusses how to get the sparse representation coefficient of the image.through this method the low resolution image can be reconstructed to high resolution image,and the resolution of the image can be increased by 1.5-2 times,and the peak signal-to-noise ratio compared to other methods can be reduced by 5%.In summary,the image obtained by traditional optical system with high resolution and large field can be improved in lightness and resolution,and this paper discusses the advantages of this method and other methods,and get the improved method for this computational imaging algorithm.