天堂国产午夜亚洲专区-少妇人妻综合久久蜜臀-国产成人户外露出视频在线-国产91传媒一区二区三区

當前位置:主頁 > 科技論文 > 電子信息論文 >

3D衍射光場的點云計算方法與加速

發(fā)布時間:2018-02-26 20:17

  本文關鍵詞: 衍射場 點云 全息圖 采樣 查找表 對稱性 出處:《安徽大學》2017年碩士論文 論文類型:學位論文


【摘要】:全息顯示能夠提供幾乎等同于真實世界的三維圖像,能充分地傳遞人類感知三維物體的所有深度刺激,將是三維成像和顯示的終極目標。一個視頻成像系統(tǒng)包括若干功能性部分,包括獲取、壓縮、傳輸和顯示等。其中,獲取是非常重要的一步。與傳統(tǒng)光全息術中采用光干涉方法相比,面向數(shù)字全息的計算產(chǎn)生全息圖的方法顯然具備對環(huán)境要求低、靈活性高、可以對物光波的數(shù)學描述進行直接編碼等優(yōu)點。但是,三維衍射場的計算涉及信號處理和光學理論等基礎性問題,仍然面臨諸多挑戰(zhàn),探尋降低三維衍射場計算復雜度的方法以及探討相應的離散化準則仍是目前的研究熱點。計算全息涵蓋極其豐富的信號處理問題,例如,Fresnel核既不是帶限也不是空間有限的,但它仍然能從它們有限的采樣中恢復完整的Fresnel衍射場。根據(jù)Wigner分布分析,信號的相干編碼比非相干編碼情況下包含更多重構所需的主要信息。因此,一方面為了產(chǎn)生生動真實的三維圖像,應該計算標量衍射形式二項級數(shù)展開的Fresnel(而不是遠場Fraunhofer)全息圖,導致產(chǎn)生的高空間帶寬積和二項級數(shù)的復波衍射場給計算全息造成巨大的挑戰(zhàn),另一方面相干的Fresnel衍射場具有一些性質(zhì)可以用作加速計算方法的基礎。計算三維衍射場的方法有很多,本文以點云方法為基礎,對Fresnel衍射場的加速計算方法進行進一步探究。首先,通過探究衍射場的采樣問題并結合菲涅耳衍射傳播在傳播方向的可逆性給出一種間接計算方法;其次,通過對查找表方法的研究并結合Gabor波帶片或Gabor透鏡的條紋圖案具有的空間對稱性,對原有查找表方法進行改進進而提高計算效率。論文的主要研究內(nèi)容與創(chuàng)新情況如下:(1)根據(jù)點云算法的采樣標準并結合菲涅爾傳播的可逆性給出雙步計算方法,使用該方法可以獨立設定源場景與全息面的采樣間距。在實際計算過程中根據(jù)需要對全息面與源場景設定不同的采樣間距,從而在保證人眼觀察效果的前提下,有效增大源場景采樣間距,減少點云數(shù)量,提高計算效率。(2)在新型查找表方法的基礎上進行改進,提出一種計算主要條紋圖案的新方法,新型查找表方法計算主要條紋圖案的過程中,一般設定為同軸全息計算,且樣本點與全息面的中心位置都在z軸,通過與全息圖的對稱性結合,則每次計算時只需計算一個對稱區(qū)域的全息圖,其它位置可以根據(jù)對稱性直接獲取。此方法可大大減少主要條紋圖案計算過程中的重復計算,從而進一步提高效率。
[Abstract]:Holographic displays can provide three-dimensional images that are almost identical to the real world, fully transmitting all the depth stimuli of human perception of three-dimensional objects. It will be the ultimate goal of 3D imaging and display. A video imaging system includes several functional components, including acquisition, compression, transmission and display. Acquisition is a very important step. Compared with the traditional optical interferometry in optical holography, the method of generating holograms based on digital holography obviously has the advantages of low environmental requirements and high flexibility. The mathematical description of object light wave can be directly coded. However, the calculation of three-dimensional diffraction field involves some fundamental problems, such as signal processing and optical theory, which still face many challenges. Exploring methods to reduce the computational complexity of three-dimensional diffraction field and discussing the corresponding discretization criteria are still the hot topics. CGH covers a wealth of signal processing problems, such as the Fresnel kernel is neither a band limit nor a limited space. But it can still recover the complete Fresnel diffraction field from their limited sampling. According to the Wigner distribution analysis, the coherent coding of the signal contains more main information needed for reconstruction than in the case of non-coherent coding. On the one hand, in order to produce vivid and real 3D images, we should calculate the Fresnel holograms (not far field Fraunhofer) holograms in the form of scalar diffraction binomial series. The resulting high spatial bandwidth product and complex wave diffraction field of binomial series pose a great challenge to CGH. On the other hand, the coherent Fresnel diffraction field has some properties which can be used as the basis of the accelerated calculation method. There are many methods to calculate the three-dimensional diffraction field, which is based on the point cloud method. The accelerated calculation method of Fresnel diffraction field is further explored. Firstly, an indirect calculation method is proposed by exploring the sampling problem of diffraction field and combining the reversibility of Fresnel diffraction propagation in the propagation direction. By studying the method of lookup table and combining the spatial symmetry of the stripe pattern of Gabor band plate or Gabor lens, The main contents and innovations of this paper are as follows: 1) according to the sampling standard of point cloud algorithm and the reversibility of Fresnel propagation, a two-step calculation method is presented. By using this method, the sampling distance between source scene and holographic plane can be set independently. In the actual calculation process, different sampling spacing between holographic surface and source scene can be set according to the need, so as to ensure the effect of human eye observation. Effectively increasing the sampling distance of source scenes, reducing the number of point clouds and improving the calculation efficiency. (2) based on the new lookup table method, a new method for calculating the main stripe patterns is proposed. In the process of calculating the main stripe patterns by the new lookup table method, the holographic calculation is generally set as coaxial holography, and the center position of the sample point and the holographic plane are both in the z axis, which combines with the symmetry of the hologram. Then only a hologram of a symmetric region is calculated for each calculation, and the other positions can be obtained directly according to symmetry. This method can greatly reduce the repeated calculation in the calculation of the main stripe patterns, thus further improving the efficiency.
【學位授予單位】:安徽大學
【學位級別】:碩士
【學位授予年份】:2017
【分類號】:O436.1;TN27

【參考文獻】

相關期刊論文 前2條

1 張成;汪東;沈川;程鴻;陳嵐;韋穗;;基于奇異值分解的可分離壓縮成像方法[J];計算機研究與發(fā)展;2016年12期

2 程鴻;陳婭萍;張成;沈川;韋穗;;基于正弦光柵調(diào)制的相位恢復[J];光子學報;2016年04期

,

本文編號:1539544

資料下載
論文發(fā)表

本文鏈接:http://sikaile.net/kejilunwen/dianzigongchenglunwen/1539544.html


Copyright(c)文論論文網(wǎng)All Rights Reserved | 網(wǎng)站地圖 |

版權申明:資料由用戶2ac0b***提供,本站僅收錄摘要或目錄,作者需要刪除請E-mail郵箱bigeng88@qq.com