Research On Computer-generated Hologram Based On Polygon-based Method

Really reproducing the 3D object field is an important development trend of display technology.The holographic 3D display retains all the information of the original light wave,so that it has the same 3D characteristics as the original object.So holographic technology is considered to be the most ideal true 3D display technology.However,although traditional holography can produce large-size high-resolution or even true-color holograms,it is difficult to get out of the laboratory because of limitation by the harsh requirements of experimental conditions and environment.Computer-generatedhologram(CGH)greatly simplifies holographic fabrication requirements,which can be displayed in the high-performance optoelectronic display devices to easily and flexibly switch between different holograms.Therefore,CGH makes true stereoscopic three-dimensional display possible.However,the heavy computational load in CGH calculation because of the huge amount of data sampling for the three-dimensional object,which is the main reason that affects the computational efficiency of the hologram for a long time.This thesis takes "accelerated holographic calculation" as the main purpose,and takes "computation holographic surface element algorithm" as the core content,and explores the method of further improving the holographic calculation rate from the aspect of algorithm.Firstly,the existing three kinds of polygon-based algorithms(traditional surface element algorithm,two-dimensional affine transformation algorithm and three-dimensional affine approximation algorithm)are deduced and summarized in detail from the principle and process.This thesis discusses the details of hologram production using different algorithm,and analyze their time-consuming characteristics,theoretically compare the advantages and disadvantages of each algorithm.Secondly,starting from the affine theory,based on the three-dimensional affine approximation algorithm,a new polygon-based method of CGH algorithm-three-dimensional affine precision algorithm-is proposed.The detailed theoretical derivation and the rigorous full-analytical expression of the light field distribution are given.In theory,the algorithm is used to overcome the limitations and time-consuming of each previous algorithm,and the correctness of the proposed algorithm is verified by simulation of a single space triangle.Thirdly,in order to objectively compare the comprehensive performance of the four polygon-based algorithms,the thesis implements the simulation calculation of the four algorithms in the same computing platform based on the MATLAB programming environment,counts the time-consuming ratio of each calculation module in each algorithm and discusses with theoretical analysis.In addition,holographic calculations are performed for multiple 3D object models of different polygon numbers.After obtaining the time consumption of each algorithm,the computational efficiency of each algorithm is comprehensively compared and discussed.Finally,the advantages and disadvantages of the four algorithms are summarized from the computational time,and it is concluded that the three-dimensional affine transformation is faster 5 times at least in rate compared to the traditional method and the 2D affine transformation algorithm.The three-dimensional affine precision algorithm proposed in this thesis is further improved by about 5 times compared with the three-dimensional affine approximation algorithm.In addition,in terms of reconstruction image quality,the three-dimensional reconstruction images of the traditional method and the two-dimensional affine transformation algorithm are significantly worse than the two three-dimensional affine transformation algorithms due to the approximation of interpolation to a certain extent.