Research On The Image Quality Assessment Of Light Field Camera

Image quality assessment provides a definite basis for the performance of the light field imaging system,and plays an important role in improving the image quality and overall performance of the system.At present,the image quality assessment theory and method for light field imaging system is not perfect,and there are still some problems to be solved.Firstly,there is an influence of the reconstruction algorithm to image quality assessment,which will introduce algorithm errors.Secondly,the spatial resolution,angle resolution information,depth information of the light field imaging system are interrelated and mutually restricted,and the current assessment methods cannot fully cover the above characteristics of the light field imaging system.Thirdly,the assessment of reconstruction algorithm of light field imaging is still lacked.Therefore,it is of great significance to establish an image quality assessment method covering the imaging characteristics of the field imaging system.In order to solve the problem of image quality assessment of field camera,based on the connection between light field and phase space optics,this dissertation studies and puts forward the assessment method of three-dimensional image quality and comprehensive image quality of light field imaging system based on the phase space theory.Through theoretical analysis,simulation and experimental test,the effectiveness of the proposed image quality assessment method is verified,and the image quality of the light field camera and the light field reconstruction algorithm are effectively evaluated.By studying the correspondence between the ray fan diagram of the optical system and ray space(the geometric representation of phase space),the ray fan diagram data of the optical system of the light field camera are extracted,corrected and processed,and the ray space distribution of the object point is obtained.The problem of image quality assessment in optical system design of light field camera is solved by taking the extension of ray space relative to straight line.A calibration method for the internal parameters and installation errors of light field camera is proposed,and the feasibility of the calibration method is verified by experiments.The main contents of this dissertation are as follows:1.Based on the relationship between four-dimensional phase space and three-dimensional coordinate space,a three-dimensional image quality assessment method for light field imaging system is proposed.The distribution of aberration of imaging system in phase space is simulated and analyzed,which proves the feasibility of phase space theory for three-dimensional image quality assessment of light field imaging system.Based on the above three-dimensional image quality assessment method of the light field imaging system,the three-dimensional image quality characteristics of the light field camera are tested experimentally,and the assessment of the reconstruction algorithm is realized.Compared with the modulation transfer function calculated by the slanted edge,the distribution of the two is consistent and the correlation coefficient is higher than 97%,which verifies the effectiveness of the proposed three-dimensional image quality assessment method.2.Based on the joint characterization of spatial information and angular spectrum information in phase space,a theoretical model for comprehensive image quality assessment of light field imaging system is established,which including the angle information,spatial information and filling factor of microlens.An image quality assessment method based on the spread volume of a point source in phase space is proposed to describe the degree of deviation from the diffraction limit of an light field imaging system.The effectiveness of the proposed image quality assessment method is verified by simulation analysis and experimental test.The signal-to-noise ratio and modulation of the reconstruction image and the corresponding subimage are tested.The test results show that the reconstruction algorithm can improve the signal-to-noise ratio of the focal plane while lose resolution information.3.The relationship between the ray fan diagram and the ray space of the optical system of the light field camera is analyzed.The ray space of the optical system of the light field camera is obtained by extracting the ray fan diagram and correcting and processing.Using the line extension of object point in ray space as an index,an image quality assessment method for the design of optical system of light field camera is proposed.Based on this assessment method,the spatial distribution of ray space and the effect on image quality of light field camera are analyzed theoretically under the condition of installation error and pupil aberration of main lens.4.The assembly error control requirements for microlens array is analyzed.The calibration method of parameters of light field camera based on the pupil distance information and the calibration method of installation error of microlens array based on angle measurement are proposed.The consistency of the same calibration parameters between the two methods is verified by experimental tests,and the feasibility of the proposed calibration method is proved.The image quality assessment method proposed in this dissertation can fully demonstrate the 3d image characteristics and multi-dimensional information characteristics of the light field imaging system.The 3d image quality assessment and comprehensive image quality assessment for the self-developed light field camera based on the proposed method is realized.