Rotating Biprism Super-resolution Wide Field Of View Imaging Technology Research

Field of View(FOV)and its resolution are key parameters in the performance evaluation of imaging systems in a wide range of machine vision applications,including remote sensing countermeasures,infrared reconnaissance and medical applications.For most conventional vision imaging systems,the simultaneous implementation of Super Resolution(SR)wide FOV imaging requires the use of high performance cameras,resulting in higher cost of use.The imaging system based on rotating dual prism scanning technology combined with a single camera not only realizes super resolution wide FOV imaging,but also has the advantages of compact structure,excellent dynamic performance,high accuracy,and wide range of scanning.The SR imaging technology uses micro-displacement adjustment of the imaging view axis to achieve sub-pixel sequence image acquisition of the specified area,combined with the SR reconstruction algorithm to achieve SR imaging of the specified area;the wide-field imaging technology uses the large angle deflection of the imaging view axis to obtain images of each area location,combined with the image stitching algorithm to achieve wide-field imaging.There are two problems in the current research based on rotating dual prism imaging technology,one is that only small-angle prism imaging experiments are conducted,and the imaging aberration problem is not considered,which hinders the development of this technology in the field of vision;the second is that for the FOV expansion,the location of the expanded region is not linked to the rotating dual prism deflection capability,which leads to insufficient FOV expansion,redundant image acquisition and other problems.To address these problems,this paper considers the imaging aberrations caused by large-angle beam deflection,and gives a solution of beam inverse tracking method image correction combined with subpixel SR reconstruction,which provides a new scheme for image acquisition location,order,and quantity in FOV extension and avoids image acquisition redundancy.The details of the work are as follows:1.the theoretical study of the rotating dual prism pointing algorithm,the multi-frame super-resolution reconstruction algorithm based on Projections Onto Convex Sets(POCS),and the optimal seam stitching algorithm.The working principle of rotating dual prism is introduced,and the relationship between the combination of prism angle and the pointing of the outgoing beam is investigated to lay the theoretical foundation for the subsequent accurate control of the imaging visual axis.2.Analysis and correction of aberrations in rotating biprism imaging.The imaging aberrations caused by the nonlinear characteristics of the rotating biprism are described,and the law of imaging aberrations is analyzed.When the view axis is located in the center of the optical axis,the imaging aberrations basically do not occur;when the farther away from the center of the optical axis,the imaging aberrations are more serious and compressed along the orientation of the view axis.According to the correction scheme of beam inverse tracking method proposed in this paper,the correction simulation experiments are conducted for different positions of the visual axis.The simulation results show that the mean change of pixel pitch is controlled below 0.5% and the fluctuation rate is reduced to within 0.6% in the corrected image compared with the aberrated image,which is a significant correction effect and lays a theoretical foundation for the image alignment technology in the subsequent SR reconstruction and image stitching.3.Research on sub-pixel scanning technology and FOV extension technology.According to the reconstruction factor,the subpixel scanning path planning was realized,and the accuracy calculation of the view axis pointing was carried out,and the alignment pixel error of the subpixel reconstruction points at four positions was less than 0.05.The position and number of reconstruction areas were planned based on the relevant parameters of the prism to achieve the maximum expansion of the FOV and avoid redundancy of data acquisition.In the FOV expansion simulation experiment and super-resolution wide-field imaging experiment,the camera’s horizontal and vertical fields of view were expanded by 2.10 times compared to each other,and the overall FOV was expanded by 4.4 times;in the super-resolution wide-field imaging experiment,in the SR reconstruction,the scene resolution was increased by 2.5 times when the reconstruction factor was 5 in the resolution plate scene reconstruction.When the reconstructed object was a book,the clarity of the text in the book was greatly improved compared with the original image;in the FOV expansion,with the building as the object,the horizontal FOV was expanded by 1.80 times and the vertical FOV was expanded by 1.67 times,and the expanded FOV was 3 times of the original camera FOV.At the same time,the clarity of the detailed parts of the scene,such as stripes,flower pots and fences,was significantly improved to achieve the expected effect.