Research On Algorithm Of Light Field Reconstruction Based On Compressed Sensing Physics

Imaging technology has been widely concerned.In recent years,with the development of virtual reality technology,light field imaging technology has become a research hotspot.This paper combines Compressed Sensing(CS)theory and computational imaging technology to physically achieve light field reconstruction with low sampling rate for real scenes.It overcomes the shortcomings of traditional multicamera light field imaging,such as large volume,high cost and difficult calibration,and solves the problem of large storage space and inconvenient transmission in the light field imaging technology based on microlens array.In this paper,the principle of compressed sensing and computational imaging technology are studied.In view of the following problems in current light field images,the angle resolution is increased by reducing the spatial resolution,and the light field image occupies a large amount of memory and is inconvenient to transmit.We reconstruct 25 light field images based on the principle of compressed sensing for a set of encoded images of a real scene.While ensuring the spatial resolution of the image,it also reduces the storage space of the light field image for easy transmission.We have improved the K-SVD algorithm for training over complete dictionary.The K-SVD algorithm is a classic algorithm for training dictionaries,which can extract and save the features of the sample set in the dictionary.In this paper,the regularization method is used to optimize the K-SVD algorithm.By introducing different norms,the problem of reducing dictionary atomic errors and overfitting is solved,and the quality of light field reconstruction is improved.This paper verifies the superiority of the improved algorithm through simulation experiments.Compared with the K-SVD algorithm,the peak signal-to-noise ratio of the reconstructed light field image has been greatly improved.The encoding acquisition process of the real scene and the reconstruction of the light field image are physically realized.First,simulation experiments are used to verify the reconstruction principle,and the experimental results prove the feasibility of the scheme.Then,an experiment based on film masks to obtain real scene coded images is carried out,which realized the code acquisition and light field reconstruction of the real scene in the spatial domain.In order to improve the controllability and adaptability of the mask,a new method of replacing the real mask by loading the electronic mask based on the spatial light modulator is proposed.It can reduce the production cost of real mask and improve the universality of physical realization process.The experimental results show that the electronic mask of the spatial light modulator can physically reconstruct the light field.