Research On Specificity Analysis And Image Restoration Of Diffraction Membrane Image

With the increasing demand of high-quality image in on-orbit surveillance,the aperture of space telescopes has also increased.However,reflecting telescopes usually face technical problems such as increasing volume and weight,difficulty in controlling and difficulty in deployying.As a new imaging technology,the diffractive membrane optical imaging system uses diffractive optical element as the primary lens of telescope,which is one of the important directions for large-aperture telescope imaging in the space.However,due to the multiple diffraction orders of the diffractive lens,the contrast of the diffractive membrane imaging system is low,and when imaging in the field experiment,noise can cause much affection.And it is more unclear than the traditional imaging system of the same aperture.Therefore,in this article,the quality assessment and specificity analysis of the diffraction membrane image are carried out,and the diffraction membrane image enhancements are carried out according to the analysis results.The main research work includes the following sections:1.The diffractive membrane imaging system is from a scientific institution,and quantitative analysis of its images are carried out from four aspects.By proposing an improved inclined edge method,the edge fitting accuracy is improved,and the MTF value is measured and analyzed;when measure the resolution,based on the Ruili criterion,an automated test plan is proposed;when measure the spatial variation of the diffraction image,by measuring the MTF value in each area by block and defining formulation to caculate the degree of spatial variation,the degree of spatial variation can be caculated;when measure the chromatic aberration of RGB imaging,this ariticle gives a quantitative calculation method of color shift.In addition,this article analyzes the imaging characteristics of the diffractive membrane imaging system from the theory of diffraction imaging and the fabrication error of diffraction imaging,and gives some qualitative analysis conclusions on the results of the above quantitative analysis.2.Propose methods of diffraction membrane image restoration.Based on the deblurring theory of maximum posterior probability and the existing deblurring framework,a blind restoration model of the diffraction membrane images is established.To improving the MTF and resolving power of the diffraction membrane images and maintaining the texture details of the diffraction images,according to the blur kernel prior of the diffraction membrane images and the clear image prior,a suitable loss function is constructed to constrain the image quality in the iterative process.At the same time,combined with the concept of fuzzy degree,the space-variant image is divided into an approximately space-invariant area,and the image is divided into blocks to restore,so as to reduce the degree of the space-variant image.3.Reduce the chromatic aberration of the diffraction membrane images.When imaging with the RGB camera,the image has serious chromatic aberration problem.By comparing with the traditional clear imaging,the difference of the edge between the diffraction membrane images and the traditional image is concluded,and the channel prior is proposed.When restore the image,G channel the with better performance can first be restorring.When restorring the image in other channels,use the G channel to constrain the restorring process of other channels to achieve the purpose of reducing the chromatic aberration.