See in 3D: state of the art of 3D display technologies

With advances in lasers, optics, and electronics, many new 3D display technologies have been proposed with prototypes in research labs or have entered the marketplace. Although some of these technologies (such as Stereoscopy) are familiar to people, other technologies, such as holography, remain far-fetched to most. This survey introduces the principles of current popular 3D display technologies, which are generally categorized into four categories: 3D movies, on-stage holograms, holographic projections and volumetric 3D displays. Furthermore, the limitations of each of the aforementioned technologies are deeply analyzed, and comparisons of these technologies are provided. Moreover, we note appropriate application situations for the various technologies. Because computer-generated hologram (CGH) technologies are considered to be the next generation of 3D display technology and have become a dominant direction in 3D display technology development, we address the challenges that CGH is currently facing and provide an insightful analysis of solutions proposed in recent years. Finally, we study the current 3D display applications associated with the four categorized technology principles.     

Dual-side see-through integral imaging 3D display system based on lens array holographic optical element

We propose a dual-side see-through integral imaging 3D display based on a lens array holographic optical element (LAHOE). The display system consists of two projectors and a LAHOE. The LAHOE is fabricated according to the theory of reflective volume holograms. An interference pattern is formed by a plane-wave beam and a spherical-wave array beam, and the interference pattern is recorded into a photopolymer material. Thus, the LAHOE is formed. When a reference beam is projected into the LAHOE, the wavefront of a convex micro-lens array is reconstructed. When a phase conjugation reference beam is projected into the LAHOE, the wavefront of a concave micro-lens array is reconstructed. So, the LAHOE possesses the optical properties of a convex micro-lens array and a concave micro-lens array. The proposed display is developed using integral imaging principle and it realizes dual-side see-through integral imaging 3D display.     

3D SSD tracking with estimated 3D planes

We present a tracking method where full camera position and orientation is tracked from intensity differences in a video sequence. The camera pose is calculated based on 3D planes, and hence does not depend on point correspondences. The plane based formulation also allows additional constraints to be naturally added, e.g., perpendicularity between walls, floor and ceiling surfaces, co-planarity of wall surfaces etc. A particular feature of our method is that the full 3D pose change is directly computed from temporal image differences without making a commitment to a particular intermediate (e.g., 2D feature) representation. We experimentally compared our method with regular 2D SSD tracking and found it more robust and stable. This is due to 3D consistency being enforced even in the low level registration of image regions. This yields better results than first computing (and hence committing to) 2D image features and then from these compute 3D pose.     

基于人体行为3D模型的2D行为识别

针对行为识别中行为者朝向变化带来的问题, 提出了一种基于人体行为3D模型的2D行为识别算法. 在学习行为分类器时, 以3D占据网格表示行为样本, 提取人体3D关节点作为描述行为的特征, 为每一类行为训练一个基于范例的隐马尔可夫模型(Exemplar-based hidden Markov model, EHMM), 同时从3D行为样本中选取若干帧作为3D关键姿势集, 这个集合是连接2D观测样本和人体3D关节点特征的桥梁. 在识别2D行为时, 2D观测样本序列可以由一个或多个非标定的摄像机采集. 首先在3D关键姿势集中为每一帧2D观测样本寻找与之最匹配的3D关键姿势帧, 之后由行为分类器对2D观测样本序列对应的3D关键姿势序列进行识别. 该算法在训练行为分类器时要进行行为者的3D重构和人体3D关节点的提取, 而在识别2D行为时不再需要进行3D重构. 通过在3个数据库上的实验, 证明该算法可以有效识别行为者在任意朝向下的行为, 并可以适应不同的行为采集环境.     

基于弹簧质点模型的二维/三维映射算法*

提出了一种用于服装设计的二维/三维映射算法.该算法基于弹簧质点变形模型,服装裁剪片二维到三维映射及三维到二维映射可以在该模型中得到统一的实现.在服装裁剪片二维到三维的映射过程中,二维裁剪片被放置在人体模型附近的初始位置,在缝合力的作用下,裁剪片自动变形并缝合到人体模型上.在服装裁剪片三维到二维映射的过程中,三维裁剪片被初始映射到指定的平面内,在弹性变形力的作用下,逐步变形并得到最终的二维裁剪片.在进行服装裁剪片二维到三维映射的过程中,考虑了干涉检验的问题.     

基于二维视图特征的三维重建

在分析现有ＣＡＤ二维参数化设计,三维参数化设计以及三维重建方法的基础上,基于工程图图形的整体宏观性,图形拓扑性及三维视图的投影规律,提出了二维视图特征的概念及一种新的工程图处理方法。     

3D sparse signal recovery via 3D orthogonal matching pursuit

Though many three-dimensional (3D) compressive sensing schemes have been proposed, recovery algorithms in most of these schemes are designed for 1D or 2D signals, which cause some serious drawbacks, e.g., huge memory usage, and high decoder complexity. This paper proposes a 3D separable operator (3DSO) which is able to completely exploit the spatial and spectral correlation to sparsify and samples the 3D signal in three dimensions. A 3D orthogonal matching pursuit (3D-OMP) algorithm is then employed to recover the 3D sparse signal, which is able to reduce the computational complexity of the decoder significantly with guaranteed accuracy. In the proposed algorithm, we represent each 3D signal as a weighted sum of 3D atoms, which allow us to sample the 3D signal with 3D separable sensing operator. Then the best matched atoms are selected to construct the 3D support set, and the 3D signal is optimally recovered from the 3D support set in the sense of the least squares. Experimental results show that the 3D-OMP approach achieves higher recovery quality but requires less computational time than the Kronecker Compressive Sensing (KCS) scheme.     

Sketch-based 3D model retrieval by incorporating 2D-3D alignment

Sketch-based 3D model retrieval is very important for applications such as 3D modeling and recognition. In this paper, a sketch-based retrieval algorithm is proposed based on a 3D model feature named View Context and 2D relative shape context matching. To enhance the accuracy of 2D sketch-3D model correspondence as well as the retrieval performance, we propose to align a 3D model with a query 2D sketch before measuring their distance. First, we efficiently select some candidate views from a set of densely sampled views of the 3D model to align the sketch and the model based on their View Context similarities. Then, we compute the more accurate relative shape context distance between the sketch and every candidate view, and regard the minimum one as the sketch-model distance. To speed up retrieval, we precompute the View Context and relative shape context features of the sample views of all the 3D models in the database. Comparative and evaluative experiments based on hand-drawn and standard line drawing sketches demonstrate the effectiveness and robustness of our approach and it significantly outperforms several latest sketch-based retrieval algorithms.     

一种2D+3D人耳融合识别方法研究

提出一种2D和3D模式相融合的人耳识别方法.利用基于Adaboost算法的人耳检测器在2D图像上进行人耳检测,在对应的深度图像中定位出人耳区域.对于2D人耳图像,利用核Fisher鉴别分析法进行特征提取,再利用最近邻分类器进行识别;对于3D人耳深度图,利用3D局部二值模式进行特征提取,结合几何约束和位置约束在测试耳和注册原型耳上进行特征点匹配,并利用匹配点数进行识别.最后将两者进行决策层融合.在UND人耳图像库上的实验结果表明,与单独2D或3D人耳识别相比,文中提出的2D+3D人耳融合识别方法在光照变化情况下能取得更好的识别性能.     

Switching directional BLUs for full‐resolution auto‐stereoscopic 3D/2D and 2D/3D LCDs

Directional backlight unit (BLU) design concept was applied to realize full‐resolution field alternate auto‐stereoscopic liquid crystal displays (LCDs) with built‐in 3D/2D and 2D/3D. The application‐oriented BLU design requires priority in realizing 2D mode or 3D mode. The switching characteristic of BLU with two confronting illuminating sets was applied to build 2D mode into 3D display and 3D mode into 2D display. An LCD with 2D mode as primary function requires higher double‐sided luminance uniformity than 3D mode. On the other hand, an LCD with 3D mode requires higher single‐sided luminance uniformity than 2D mode. For increasing the double‐sided luminance uniformity for 2D mode as a primary function of the display, the former BLU design was realized by using uniform prismatic structure, that is, using uniform bilateral prismatic structure with a cross section of isosceles triangle on the back surface of the light‐guide plate. For increasing single‐sided luminance uniformity for 3D mode as a primary function of the display, the latter was realized by using graded angle design, that is, unilateral prisms with a cross section of a scalene triangle on the back surface of the light‐guide plate. The LCDs of small‐sized‐handheld game devices and still cameras were fabricated using the 3D/2D and 2D/3D displays.     

3D Euclidean versus 2D non-Euclidean: two approaches to 3D recoveryfrom images

Methods of 3D recovery in computer vision for computing the shape and motion of an object from projected images when an object model is available are classified into two types: the 3D Euclidean approach, which is based on geometrical constraints in 3D Euclidean space, and the 2D non-Euclidean space. Implications of these two approaches are discussed, and some illustrating examples are presented     

二维概念格图形的三维重构研究

介绍了二维概念格图形向三维空间转化和延伸的必要性和现状。通过对传统概念格图形分层定位布局方法的研究与分析,提出并实现了一种新的以具有大量的平行四边形和有向线段为基本特征的概念格在三维空间的自动布局算法,描述了一种基于该算法的二维概念格图形的三维重构机制,有效地解决了节点横向过度扩张的问题并减少了线段交叉,较好地实现了复杂概念格图形的三维可视化,为知识发现和知识处理提供了良好的基础。     

3D extension to GKS

The Graphical Kernel System GKS has [1] been established as the first standard in the field of Computer Graphics covering two-dimensional (2D) graphics. Now work is going on to develop standards in related areas. One important effort is the extension of GKS for three-dimensional (3D) graphics. This paper will briefly overview the history of standardization efforts with respect to 3D graphics and then report the current activities of various national and international standardization bodies for extending GKS to 3D. Then the paper will concentrate on GKS-3D [2], a proposal for a 3D extension of GKS which is developed by the Dutch standardization committee NNI in close collaboration with the International Organization for Standardization ISO/TC97/SC21/WG2. Technical work is expected to finish in 1985. Scope and purpose of this future 3D standard and goals of the design are given and the functionality of the 3D extension is described in some detail. As technical work on GKS-3D is going on, changes may occur to the standard document. The major issues will be surveyed and trends will be sketched.     

An evaluation of multimodal 2D+3D face biometrics

We report on the largest experimental study to date in multimodal 2D+3D face recognition, involving 198 persons in the gallery and either 198 or 670 time-lapse probe images. PCA-based methods are used separately for each modality and match scores in the separate face spaces are combined for multimodal recognition. Major conclusions are: 1) 2D and 3D have similar recognition performance when considered individually, 2) combining 2D and 3D results using a simple weighting scheme outperforms either 2D or 3D alone, 3) combining results from two or more 2D images using a similar weighting scheme also outperforms a single 2D image, and 4) combined 2D+3D outperforms the multi-image 2D result. This is the first (so far, only) work to present such an experimental control to substantiate multimodal performance improvement.     

Taking a 2D educational title into 3D

Maturing software for creating real-time interactive 3D (3D/VR) and hardware enabling its delivery give us the chance to explore practical issues involved in making commercially viable 3D/VR content. In October 1996, The Learning Company released the educational title Logic Quest, built using Sense8's WorldToolKit software. One of the first commercial 3D/VR children's titles for Windows 95, Logic Quest provides an example of what is involved in converting a 2D title (Think Quick!) to 3D. Studying the process the company followed lets us evaluate some of the challenges and benefits of moving a 2D title to 3D     

基于相关型图像传感器3维人脸成像的3维AAMs人脸识别方法的研究

主动外观模型是基于统计分析建立物体2维模型的有效方法,它融合了目标的形状和纹理信息。在基于相关型图像传感器3维人脸成像的基础上,提出了一种建立3维人脸模型的方法,该方法利用由相关型图像传感器得到的深度信息和与之对应的亮度信息将2维AAMs扩展为3维AAMs,融合人脸的形状,纹理和深度信息来构建3维人脸模型。人脸识别实验结果表明,该方法在不同人脸姿态,表情和光照条件下识别效果要优于Eigenface和2维AAMs。     

二维Gabor特征与三维NP-3DHOG特征融合的人脸识别算法

基于二维图像的人脸识别算法提取人脸纹理特征进行识别,但是光照、表情、人脸姿态等会对其产生不利影响。三维人脸特征能更精确地描述人脸的几何结构,并且不易受化妆和光照的影响,但只采用三维人脸数据进行人脸识别又缺少人脸纹理信息,因此文中将二维人脸特征与三维人脸特征相融合进行人脸识别。采用基于Gabor变换的二维特征与基于新的分块策略的三维梯度直方图特征相融合的算法进行人脸识别。首先,提取二维人脸的Gabor特征;然后,提取三维人脸基于新的分块策略的三维梯度直方图特征,旨在提取人脸的可辨别性特征;接下来,对二维人脸特征与三维人脸特征分别使用线性判别分析子空间算法进行训练,并使用加法原则融合两种特征的相似度矩阵;最后,输出识别结果。     

Reactive 2D/3D garment pattern design modification

This paper presents a new 3D garment simulation result update algorithm for the 2D garment pattern design modification. The proposed algorithm enables the 3D garment fitting simulation result directly to react to the modification in the 2D patterns. The algorithm performs a topological invariant deformation of the 2D pattern mesh after the boundary of the 2D pattern undergoes a topological consistent modification. The length of each of the edges in the mesh defined as the equilibrium state parameter is updated and then directly used in the 3D garment fitting simulation to update the original simulation result. The advantage of the proposed algorithm is that the mesh topology of the 2D garment pattern is preserved and thus simplifies the numerical scheme by maintaining the consistency of the matrix equation. With this approach, the 3D garment fitting simulation does not need to repeat the entire simulation for every modification and can react to the 2D pattern modification efficiently and speedily.     

Java 3D中的Text2D的扩展与应用

Java3D中Text2D类虽然能生成二维文本,但由于文本的字体、字号、颜色在文本生成后就不可改变,字的排列也只有“从左向右”方式。所以,Text2D不能满足交互式建立虚拟环境中二维文本的需要。Java3D中的Text3D类支持交互式地建立文本,但Text3D的三维文本却不能退化为二维文本,这使得Text3D不能替代Text2D。所以,有必要扩展Text2D类以支持交互地建立虚拟环境中二维文本。文章给出实现这一扩展的关键技术,为用Java3D开发交互式虚拟环境建模器的二维文本生成功能提供支持。扩展的Text2D在智能型虚拟汽车驾驶道路环境建模器中的使用表明它是有效的。