Research On Computational Imaging Of Cassegrain-type Optical Systems With Large Field Of View

The Cassegrain-type optical system can realize long focal length and large aperture imaging,and is mainly applied to aviation,aerospace,astronomy and other long-distance detection fields.The classic Cassegrain-type optical system can achieve perfect imaging of the on-axis point,but the off-axis aberration increases rapidly after leaving the on-axis point,which seriously affects the imaging quality and limits the imaging field of view.As a derivative structure of the Cassegrain-type system,the R-C system has an enlarged field of view,but from an application point of view,its field of view is still small.In order to better meet the requirements of a larger field of view,a correction mirror or a correction lens group can be added to the front of the behind of the two mirrors to correct the aberration of the off-axis field of view,but this method makes the structure complicated and is not conducive to the system compactness and lightweight.In this dissertation,the computational imaging technology of Cassegrain-type optical system with large field of view is researched.Combining the degradation model of optical system,a design method of optical joint image restoration is proposed,which can expand the field of view of Cassegrain-type optical system with only primary and secondary mirrors.Firstly,the aberration characteristics of Cassegrain-type optical system and R-C optical system are researched in this dissertation.Considering the requirements of front-end optical design and back-end restoration processing,the design method of preserving coma and field curvature is proposed,and the Cassegrain-type computational imaging optical system is designed.Then,the wavefront aberration of each sampling field of view of Cassegrain-type computational imaging optical system is fitted by using Zernike polynomial,wavefront aberration model which changes continuously with the field of view is built according to the fitting data and aberration variation law,the fast building of the continuous point spread function model in the full field of view is realized by establishing the generalized pupil function and fourier transform.Finally,the image blur in a large field of view is eliminated to a certain extent by applying the spatial variation deconvolution algorithm according to the combination of the point spread function model and the image of Cassegrain-type computational imaging optical system in a large field of view.In order to verify the correctness and effectiveness of the proposed method.In this dissertation,we first design a Cassegrain-type optical system whose F number is 5.5,band is 0.9 ? 1.7 ?m,f is 470 mm,and quickly build the point spread function model.Then the Cassegrain-type optical system is simulated and the imaging results are processed.Finally,the final imaging results are evaluated.The final MTF of this method is 0.48 at 33 lp / mm,0.23 ° field of view,which is about 0.28 higher than that of the unprocessed image;0.30 at0.47 ° field of view,which is about 0.21 higher than that of the unprocessed image;0.28 at0.75 ° field of view,which is about 0.17 higher than that of the unprocessed image.After processing,the MTF of off-axis field of view is significantly improved,and the image quality is significantly improved.The results show that this method can realize clear imaging of Cassegrain-type optical system with a large field of view,which provides a new method for Cassegrain-type optical system imaging with a large field of view.