光电工程
光電工程
광전공정
OPTO-ELECTRONIC ENGINEERING
2014年
1期
48-53
,共6页
计算机全息%并行计算%冗余光波%二步算法%子全息
計算機全息%併行計算%冗餘光波%二步算法%子全息
계산궤전식%병행계산%용여광파%이보산법%자전식
CGH%parallel computing%redundant lightwave%two-step algorithm%sub-hologram
全息图的生成速度影响了全息三维显示的实用化,大尺寸、大场景全息图的生成尤为困难,为了解决此问题,将人眼跟踪技术和指向光技术用于全息显示,形成主动式全息显示是一种解决动态全息显示的途径。针对主动式全息显示的需要,提出一种基于二步全息算法及空间冗余光波去除的并行全息图生成方案。首先根据二步全息算法计算行列贡献分量,结合空间冗余光波去除原理确定子全息图范围,合理设计基于GPU的CUDA并行计算方案,以实现大尺寸、大场景全息图的快速生成。实验表明这种方法有效可行,二步算法的引入使计算速度在并行计算的基础上再提高10倍左右,空间冗余光波的去除有效克服大场景与空间采样间隔之间的矛盾。
全息圖的生成速度影響瞭全息三維顯示的實用化,大呎吋、大場景全息圖的生成尤為睏難,為瞭解決此問題,將人眼跟蹤技術和指嚮光技術用于全息顯示,形成主動式全息顯示是一種解決動態全息顯示的途徑。針對主動式全息顯示的需要,提齣一種基于二步全息算法及空間冗餘光波去除的併行全息圖生成方案。首先根據二步全息算法計算行列貢獻分量,結閤空間冗餘光波去除原理確定子全息圖範圍,閤理設計基于GPU的CUDA併行計算方案,以實現大呎吋、大場景全息圖的快速生成。實驗錶明這種方法有效可行,二步算法的引入使計算速度在併行計算的基礎上再提高10倍左右,空間冗餘光波的去除有效剋服大場景與空間採樣間隔之間的矛盾。
전식도적생성속도영향료전식삼유현시적실용화,대척촌、대장경전식도적생성우위곤난,위료해결차문제,장인안근종기술화지향광기술용우전식현시,형성주동식전식현시시일충해결동태전식현시적도경。침대주동식전식현시적수요,제출일충기우이보전식산법급공간용여광파거제적병행전식도생성방안。수선근거이보전식산법계산행렬공헌분량,결합공간용여광파거제원리학정자전식도범위,합리설계기우GPU적CUDA병행계산방안,이실현대척촌、대장경전식도적쾌속생성。실험표명저충방법유효가행,이보산법적인입사계산속도재병행계산적기출상재제고10배좌우,공간용여광파적거제유효극복대장경여공간채양간격지간적모순。
In holographic three-dimensional display, the generation speed of the hologram affected its practical application. Particularly, it is difficult to generate hologram of the large size and large scene fast. In order to solve the problem, the pupil tracking and directional backlight were applied to holographic display. The formation of active holographic display is a way to solve the dynamic holographic display problem. For the need of active holographic display, a parallel hologram generation scheme based on two-step hologram algorithm and spatial redundancy lightwave removal algorithm was proposed. Firstly, rank contribution components were calculated by the two-step algorithm. Secondly, the scope of the sub-hologram was determined by the spatial redundancy lightwave removal algorithm. Finally, a reasonable CUDA parallel computing program was designed to generate hologram of the large size and large scene fast. The experimental results indicated that the algorithm was effective and feasible. With the two-step algorithm, the computing speed was accelerated about ten times on the basis of parallel computing technology. The contradiction between the large scene and spatial sampling interval was also overcome effectively by the spatial redundancy lightwave removal algorithm.