基于线性结构的几种特殊矩阵的压缩存储

数据压缩技术是计算机软件领域中一个重要的研究课题,在图像、视频、音频等多媒体信息处理方面都需要进行数据压缩存储,而矩阵又是工程设计中一个常用的数学对象。以数组作为存储基础,采用Java编程语言,实现三角矩阵、对称矩阵、稀疏矩阵等特殊矩阵的压缩存储算法,抛砖引玉、以飨读者。     

基于霍特林变换的三维医学图像快速配准算法

提出了一种新的基于霍特林变换的三维医学图像快速配准算法,这是将数据压缩技术用于图像配准的一种创新性尝试。传统的基于灰度的方法需要考虑整个三维数据的灰度信息,计算复杂度大,无法满足临床需要。论文将Otus算法与互信息量技术相结合提出了一种新的图像分割算法,用于提取待配准物体,从而得到物体的向量表示;然后通过霍特林变换的平移和旋转性质完成配准。实验结果表明此方法能准确,快速地处理图像刚性配准问题,特别适用于三维医学图像的配准。     

基于SFM与深度学习融合的水果体积测量算法

以生活常见的水果作为研究对象，结合SFM对图像序列的深度估计以及神经网络重建三维结构的优点，提出SFM算法融合深度学习三维重建的水果体积测量算法。对单目相机采集的水果多视角图像进行研究，分析图像重建以及估计大小的方法，搭建快速、便捷估计水果实际体积算法框架。使用神经网络快速推理水果结构，解决三维重建构建稠密点云耗时长的缺点，利用多视角图像获取稀疏点云，估计目标尺寸，提高采样的便利性。实验结果表明，该算法能快速重建水果三维模型，实现简单、快速、较精确的水果体积测量。     

MR码编码与译码算法研究及仿真实验

传真图像的信息量非常大,对其进行数据压缩,不仅可以节省存储空间,而且可以有效地降低传输时间。针对MR码,设计了一种编码和译码算法,并利用MATLAB软件进行实现;采用该算法对传真图像进行数据压缩,并比较MH编码和MR编码的压缩比,仿真实验表明,该算法可以满足传真图像对压缩比和压缩速度的要求,并证明了该算法的可行性。     

基于差分算法和Kinect的数媒交互系统设计

为了有效实现数媒交互,提升三维模型重建速度、精度及数媒交互效果,设计了一种基于差分算法和Kinect的数媒交互系统。首先读取、转换与存储不同格式的图像数据,利用视觉表达灵敏度差分算法预处理图像数据,利用面绘制技术重建三维模型,再使用Kinect采集视野内物体深度值,选取与Kinect距离最近的用户当作第一操作用户,通过骨骼追踪数据捕捉用户姿势,定义骨骼数据,编辑动作指令,依据键盘映射完成Unity3D和Kinect的数据通讯,实现数媒交互。实验证明：设计系统的三维模型重建速度快、重建精度高,数媒交互展示效果更好。     

基于数据结构化的三维动画压缩方法研究

在三维动画应用日益广泛的背景下,研究三维动画数据压缩方法对其数据的存储、传播和渲染 技术至关重要。基于图像数据的结构化特征,计算机视觉领域提出了一系列高效图像压缩技术。鉴于三维动画 数据与图像数据结构的差异,探讨三维动画数据结构化方法,进而探索改进图像算法以实现一种新的三维动画 数据压缩方法。具体而言,首先研究对三维模型顶点序列进行优化调整;进而基于图像压缩算法的模块研究动 画数据的子块尺寸优化以及图像压缩算法量化矩阵优化方法;最后,通过采用不同类型动画数据进行测试,实 验结果表明优化图像压缩算法在压缩比和时效均得到有效提升和保障。     

若干细粒度三维交互算法的实现

在三维软件中涉及到大量三维交互操作,尤其是三维编辑软件中.三维交互的好坏直接影响用户的操作感.给出一种三维交互算法,此算法可以沿物体自身坐标系的原点所在的面、坐标面、三个坐标轴拖动物体和绕坐标轴旋转这个物体.算法操作感觉良好,实现了在拖动的过程中物体紧跟随鼠标的效果.     

NTT--数论变换算法在图像压缩技术中的应用研究

提出了一种全新的图像数据压缩算法,即数论变换（Ｎｕｍｂｅｒ Ｔｈｅｏｒｙ Ｔｒａｎｓｆｏｒｍａｔｉｏｎ,ＮＴＴ）算法,证明了在以正整数ｐ为模的整数环Ｚｐ上ＮＴＴ是线性正交变换,以及在Ｚｐ上具有卷积特性等,设计了具有ＦＦＴ类型的快速算法,该算法可采用移位操作实现,其程度优于ＤＣＴ变换,最后通过实例比较,说明了该算法在图像数据压缩中表现出运算速度快、精度高和压缩效果好等优点,ＮＴＴ算法的研究,为图像压     

基于全景深度图像的汽车外形重建方法

将全景深度图像与虚拟现实建模语言VRML的特点有机结合起来,实现了由深度图像到三维模型的快速重建及车身外形的多细节层次描述。该算法在网格均匀细化基础上,结合图像处理方法,能有效处理车身外形的孔洞部分,且较准确地反映车身外形。利用所得到的三维VRML模型,可方便地进行模型交互修改,并实时显示;在汽车外形概念设计和用户个性化外形定制系统的应用中,对VRML模型的建立提出了一种新方法。     

基于全景深度图像的汽车外形重建方法

将全景深度图像与虚拟现实建模语言VRML的特点有机结合起来,实现了由深度图像到三维模型的快速重建及车身外形的多细节层次描述.该算法在网格均匀细化基础上,结合图像处理方法,能有效处理车身外形的孔洞部分,且较准确地反映车身外形.利用所得到的三维VRML模型,可方便地进行模型交互修改,并实时显示;在汽车外形概念设计和用户个性化外形定制系统的应用中,对VRML模型的建立提出了一种新方法.     

Localization and comparison of two free-form surfaces

Comparison of two free-form surfaces based on discrete data points is of paramount importance for reverse engineering. It can be used to assess the accuracy of the reconstructed surfaces and to quantify the difference between two such surfaces. The entire process involves three main steps: data acquisition, 3D feature localization and quantitative comparison. This paper presents models and algorithms for 3D feature localization and quantitative comparison. Complex free-form surfaces are represented by bicubic parametric spline surfaces using discrete points. A simple yet effective pseudoinverse algorithm was developed and implemented for localization. It consists of two iterative operations, namely, constructing a pseudo transformation matrix and point matching. A computing algorithm was developed to compare two such surfaces using optimization techniques. Since this approach does not involve solving non-linear equations for the parameters of positions and orientations, it is fast and robust. The algorithm was implemented and tested with several examples. It is effective and can be used in industry for sculptured surface comparison.     

Design of a three-dimensional near real-time image processing system

The airborne three-dimensional (3D) imager, which integrates global positioning system (GPS), attitude measurement unit (AMU), scanning laser rangefinder (SLR) and spectral scanner, has been developed successfully. The spectral scanner and SLR use the same optical system, which ensures that the laser point matches pixel seamlessly. The distinctive advantage of the airborne 3D imager is that it can produce geo-referenced imagery and DSM (digital surface models) without any ground control points (GCPs). It is no longer necessary to obtain GCPs, and the software can process the data to produce DSM and a geo-referenced image in near real-time, so the airborne 3D imager is many times faster than the traditional approaches. The processing procedure of the software involves decomposing and checking the raw data, processing GPS data, calculating the positions of laser sample points, producing a geo-referenced image, producing DSM, and mosaicking strips. In this paper, we introduce the principle of the airborne 3D imager first and then focus on the fast processing algorithm. The flight tests and processed results demonstrate that the processing technique is feasible, and can meet the requirement of near real-time applications.     

An Efficient Parallel Algorithm for Computing the Gaussian Convolution of Multi-dimensional Image Data

In this paper, we propose a parallel convolution algorithm for estimating the partial derivatives of 2D and 3D images on distributed-memory MIMD architectures. Exploiting the separable characteristics of the Gaussian filter, the proposed algorithm consists of multiple phases such that each phase corresponds to a separated filter. Furthermore, it exploits both the task and data parallelism, and reduces communication through data redistribution. We have implemented the proposed algorithm on the Intel Paragon and obtained a substantial speedup using more than 100 processors. The performance of the algorithm is also evaluated analytically. The analytical results confirming with the experimental results indicate that the proposed algorithm scales very well with the problem size and number of processors. We have also applied our algorithm to the design and implementation of an efficient parallel scheme for the 3D surface tracking process. Although our focus is on 3D image data, the algorithm is also applicable to 2D image data, and can be useful for a myriad of important applications including medical imaging, magnetic resonance imaging, ultrasonic imagery, scientific visualization, and image sequence analysis.     

The Fast Evaluation Strategy for Evolvable Hardware

An evolutionary algorithm implemented in hardware is expected to operate much faster than the equivalent software implementation. However, this may not be true for slow fitness evaluation applications. This paper introduces a fast evolutionary algorithm (FEA) that does not evaluate all new individuals, thus operating faster for slow fitness evaluation applications. Results of a hardware implementation of this algorithm are presented that show the real time advantages of such systems for slow fitness evaluation applications. Results are presented for six optimisation functions and for image compression hardware.     

球面模型的逼近与数据压缩及推广

提出了一种三维物体表面的逼近表示与数据压缩方法.该方法可以在不增加表示物体表面的数据量(例如面片数量)的情况下,使逼近误差降低约1/2;在逼近误差不变的情况下,使表示物体表面的数据量大幅下降.提出了用与最基本的三维物体--球体表面相交的面片来表示球面的方法,将该方法扩展到了一般曲面.理论分析与实验数据表明,新算法与传统方法相比,其数据压缩比约为35%.该研究在虚拟现实技术和三维模型的数据压缩及传输等领域有较重要的学术及应用价值.     

Monocular Template-based Reconstruction of Inextensible Surfaces

We present a monocular 3D reconstruction algorithm for inextensible deformable surfaces. It uses point correspondences between a single image of the deformed surface taken by a camera with known intrinsic parameters and a template. The main assumption we make is that the surface shape as seen in the template is known. Since the surface is inextensible, its deformations are isometric to the template. We exploit the distance preservation constraints to recover the 3D surface shape as seen in the image. Though the distance preservation constraints have already been investigated in the literature, we propose a new way to handle them. Spatial smoothness priors are easily incorporated, as well as temporal smoothness priors in the case of reconstruction from a video. The reconstruction can be used for 3D augmented reality purposes thanks to a fast implementation. We report results on synthetic and real data. Some of them are compared to stereo-based 3D reconstructions to demonstrate the efficiency of our method.     

基于KD-tree剖分的三维动态场景快速有效压缩

为充分利用GPU并行计算特点,实现对三维动态数据的快速有效压缩,降低网络带宽的限制,提出一种基于KD-tree剖分的快速有效压缩方法。首先使用KD-tree在第0帧对整个三维场景进行划分,并对每个叶子节点进行刚体的并行构造;建立能构造刚体的叶子节点和均匀划分的三维网格之间的映射关系,在三维空间使用并查集合并并行构造的刚体;最后将压缩后的动态数据传输到客户端并重构一定时间内的三维动态场景。算法可以极大提高服务器端数据的压缩速度,有效减少需要传输的数据量。实验结果表明：该算法在保证压缩质量的同时,可以对原始三维动态场景进行快速有效压缩,有效降低网络带宽对数据传输的限制。     

并行DSP小波图象压缩的实现研究

小波图象压缩逄法是基于多分辨率分析上的一种压缩方法,它能够根据人们的视觉效应对图象信息进行针对性的处理,达到高效压缩图象的目的。小波变换算法的工程实现需要高性能处理器的支持,并行DSP（Digital Signal Processor)处理器TMS320C80就具有强大的图象处理功能,研究了在并行DSPTMS320C80上实现小波变换图象压缩的快速算法、优化程序设计及并行处理方式,并加以实现。     

基于EZW的牙颌模型三维数据压缩

针对牙颌模型深度数据的压缩处理问题,阐述了嵌入零树小波(Embedded Zerotree Wavelet,EZW)编码算法的基本原理和算法实现,讨论了基于感兴趣区域的EZW算法及其实现,并在实验室进行了深度数据和三维重构图像的基于感兴趣区域的EZW编码实验.实验表明,EZW编码既能实现图像的有损压缩也可实现图像的无损压缩,并具有很大的压缩比.