节点文献

固态体积式真三维显示关键技术研究与实现

Research and Implementation on Key Technologies of Solid-state 3D Volumetric Display

【作者】 方勇

【导师】 吴华夏; 吕国强;

【作者基本信息】 合肥工业大学 , 计算机应用技术, 2017, 博士

【摘要】 体显示是目前实现真三维显示的主流技术路径之一,而静态体积式显示无任何移动部件,具有非常高的发展潜力。本文阐述了真三维显示的视觉机理和显示特性,研究并实现了海量显示数据的高速实时处理方法;研究了深度抗锯齿、灰度级修正、闪烁评价及优化等真三维成像质量的提升与优化方法;完成了真三维显示原型样机的研制。本文的主要研究内容包括:1)真三维显示的视觉机理与成像分析。讨论了视觉暂留、似动现象及人眼的空间与时间分辨率,这些视觉特性是形成真三维显示的基础。阐述了体素、显示空间及真三维成像的原理和特点,提出了“体柱”的概念。利用全光函数分析研究了真三维显示的数据流量。相关视觉机理和成像基础的研究,为真三维显示系统的研究与设计提供了理论依据。2)海量显示数据的高速实时处理方法研究与实现。分析了真三维显示数据的处理流程,提出了体数据的生成和编码传输方法。阐述了图像预处理方法、图像引擎的设计目标和功能描述。深入研究了图像引擎数据处理过程中的数据格式变换及地址映射方法,建立了相关的数学模型。在理论研究的基础上提出了真三维图像引擎的FPGA实现方法,实现了海量显示数据的高速实时处理。3)真三维成像质量提升和优化方法研究。针对真三维显示中由于显示体结构和光电特性等引起的各种显示缺陷,在理论研究和实验验证的基础上提出了一系列显示质量改善方法。针对深度不连续问题,提出了一种基于分类映射的深度抗锯齿算法。提出了层间梯度的概念,分析了不同映射类型对于深度连续性的影响,给出了体素映射类型的判定算法。针对灰度级失真问题,在分析显示体光电特性的基础上推导出成像亮度变化函数及灰度级偏差。通过调整显示体驱动时序及重新分配子帧,结合反伽马校正,实现了灰度级的修正。针对真三维显示的闪烁优化问题,提出了一种时间对比敏感度函数加权幅值谱对比度的闪烁程度评价模型及一种层间交织扫描的显示体刷新方法,完成了两组视觉感知实验,并通过统计分析验证了评估模型的准确性和刷新方法的有效性。4)真三维显示系统原型样机的研制与实现。提出了真三维显示系统的整体架构方案,详细论述了显示数据处理电路、投影光路、显示体及驱动等子系统的设计与装配。重点阐述了DLP控制电路的数据接收和数字微镜器件的复位时序设计、光路结构和光源设计、显示体光电特性和驱动方法设计。对系统的主要性能参数进行了测试和评价,并进行了三维成像实验。结果表明,本文所研制的真三维显示原型样机整体性能达到国际先进水平。本文所研制的真三维显示原型样机性能达到预期设计目标,同时在研制过程中掌握了具有自主知识产权的核心关键技术,为体积式真三维显示技术的理论研究和工程实践积累了经验并提供了重要的实物验证平台。

【Abstract】 Volumetric display is one of the main technologies to achieve true three-dimensional display,and the static volumetric display is characteristic of no moving parts,which has a very high development potential.In this dissertation,the visual mechanism and imaging mechanism of the true three-dimensional display are expounded,and the high speed real-time processing method of massive display data and its FPGA implementation are deeply studied.The true 3D imaging quality improvement and optimization methods are proposed,including depth anti-aliasing algorithm based on classified mapping,gray-scale correction method and optimization and evaluation method of flicker.Finally,the research and development of the true 3D display prototype are completed.The main contributions of the dissertation are as follows:1)Visual mechanism and imaging analysis of true 3D display.The vision persistence,apparent movement phenomenon and spatial and temporal resolution of human sight,which form the basis of volumetric display,are discussed.Voxel,display space and the principle and characteristics of true three-dimensional imaging are expounded,and the concept of "volumetric column" is proposed.The data flow of the true three-dimensional display is studied by using Plenoptic function.The study of the related visual mechanism and the basis of imaging provides a theoretical basis for the research and design of true 3D display system.2)Research and implementation of high-speed real-time processing method for massive display data.The processing flow of true 3D display data is analyzed,and the generation,encoding and transmission methods of true 3D volumetric data are presented.The image preprocessing method and the design goal and function description of the image engine are expounded.Various data conversion and address mapping methods in image engine processing are studied in depth,and related mathematical models are established.On the basis of theoretical research,a FPGA implementation method of the true 3D image engine is proposed,which can realize the high-speed and real-time processing of massive display data.3)Research on the method of improving the quality of true 3D imaging.In view of the display defects caused by structure and photoelectric characteristics of the display body,a series of display quality improvement methods are put forward.Aiming at the problem of depth discontinuity,a true 3D depth antialiasing algorithm based on classified mapping is proposed.The concept of inter layer gradient is proposed,and the influence of different mapping types on depth continuity is analyzed.Besides,an algorithm for determining the type of voxel mapping is given.In order to deal with the problem of gray-scale distortion,the imaging luminance variation function and gray-scale deviation are derived by analyzing the photoelectric characteristics of the display volume.By adjusting the driving sequence of the display volume and the sub-frame length,combined with gamma correction,the gray-scale correction is realized and the display contrast is improved.In order to evaluate flicker degree of true 3D display,a method of temporal contrast sensitivity function weighted amplitude contrast and a display volume refreshing method of inter-layer scanning are proposed,and two groups of visual perception experiment are completed.By means of the correlation analysis between the objective quantification results of the evaluation model and the scoring results of the subjective experiments,the accuracy of the evaluation model and the validity of the refreshing method are verified.4)Development and implementation of a prototype of true 3D system.The overall architecture of true 3D display system is proposed,and design and assembly of subsystems such as display data processing circuit,projection light path and display volume and its driving are discussed.The data receiving of DLP circuit and DMD reset timing design,the light path structure and light source design and the display volume photoelectric characteristics and driving method design are expounded.The main performance parameters of the system are tested and evaluated,and the three-dimensional imaging experiments are carried out.The results show that the overall performance of the prototype developed in this dissertation can reach the international advanced level.The performance of the prototype of true 3D display system developed in this dissertation achieves the desired design goals.In the process of development,the key technologies of independent intellectual property rights has been mastered.The set up of this prototype accumulates experience and provides an important physical verification platform for future theoretical research and engineering practice of volumetric true 3D display technology.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络