Design And Implementation Of Integral Imaging System Based On Sparse Camera Array

Integral imaging is stereo display technology,which can record the complete scene information and realize the reconstruction process through optical devices.The technology has the full parallax,true color,visual fatigue-free viewing and a simple structure,which has gradually become one of the most promising 3D display technologies.The integral imaging system can be divided into two parts:pickup and display stage.In the pickup stage,the 3D information of the object is recorded on the elemental image array by lens array recording.In the display stage,the 3D image can be reproduced by optical or computational reconstruction.The use of sparse camera arrays to realize the pickup and display process of the integral imaging system can simplify the complexity of the system,improve the resolution and reduce the cost.However,this method still has room for further improvement in many aspects.(1)fewer number of picked-up view,the multi-view synthetic efficiency is low,and the quality of the synthesized view to be improved.(2)The display resolution still needs to be further improved to meet the actual needs of society.An integral imaging system based on two-dimensional view point synthesis with hierarchical clustering is proposed to improve the efficiency of view point collection,reduce the cost of view point collection,and ensure the quality of view point synthesis.In this integral imaging system,a sparse viewpoint image is collected using a motorized translation stage combined with a single camera viewpoint collection method.In order to improve the efficiency of viewpoint collection,the two-dimensional virtual viewpoint synthesis method based on hierarchical clustering is used to realize the virtual viewpoint synthesis between images;the elemental image array is obtained through pixel mapping to realize the computational reconstruction process.Experimental results show that the system can not only meet the high-efficiency three-dimensional information collection,but also improve the resolution quality and increase the application flexibility while ensuring quality of the synthetic viewpoint.In order to improve the display resolution,a super-resolution integral imaging system based on viewpoint synthesis is proposed.In order to verify the feasibility of the system,the integral imaging virtual scene experiment and real scene experiment are designed to verify the proposed scheme.In the virtual scene experiment,3Dmax software is used to build a camera array and a three-dimensional scene model to complete the acquisition process of the elemental image array;super-resolution technology is used to process the elemental image to obtain a high-resolution elemental image array and realize the computation reconstruction process.In the process of real scene acquisition experiment,the sparse view point image array is used to synthesize the multi-view point image array,and then the image resolution is improved through the super-resolution algorithm.Experimental results show that the system via the view synthesis and super-resolution method to achieve improved display quality.