Bottom-up pyramid cellular acceptors with three-dimensional layers

In 1997, C.R. Dyer and A. Rosenfeld introduced an acceptor on a two-dimensional pattern (or tape), called the pyramid cellular acceptor, and demonstrated that many useful recognition tasks are executed by pyramid cellular acceptors in time proportional to the logarithm of the diameter of the input. They also introduced a bottom-up pyramid cellular acceptor which is a restricted version of the pyramid cellular acceptor, and proposed some interesting open problems of the bottom-up pyramid cellular acceptors. On the other hand, we think that the study of threedimensional automata has been meaningful as the computational model of three-dimensional information processing such as computer vision, robotics, and so forth. In this paper, we investigate about bottom-up pyramid cellular acceptors with three-dimensional layers, and show their some accepting powers. This work was presented in part at the 13th International Symposium on Artificial Life and Robotics, Oita, Japan, January 31–February 2, 2008     

Perceptual Scale-Space and Its Applications

When an image is viewed at varying resolutions, it is known to create discrete perceptual jumps or transitions amid the continuous intensity changes. In this paper, we study a perceptual scale-space theory which differs from the traditional image scale-space theory in two aspects. (i) In representation, the perceptual scale-space adopts a full generative model. From a Gaussian pyramid it computes a sketch pyramid where each layer is a primal sketch representation (Guo et al. in Comput. Vis. Image Underst. 106(1):5–19, 2007)—an attribute graph whose elements are image primitives for the image structures. Each primal sketch graph generates the image in the Gaussian pyramid, and the changes between the primal sketch graphs in adjacent layers are represented by a set of basic and composite graph operators to account for the perceptual transitions. (ii) In computation, the sketch pyramid and graph operators are inferred, as hidden variables, from the images through Bayesian inference by stochastic algorithm, in contrast to the deterministic transforms or feature extraction, such as computing zero-crossings, extremal points, and inflection points in the image scale-space. Studying the perceptual transitions under the Bayesian framework makes it convenient to use the statistical modeling and learning tools for (a) modeling the Gestalt properties of the sketch graph, such as continuity and parallelism etc; (b) learning the most frequent graph operators, i.e. perceptual transitions, in image scaling; and (c) learning the prior probabilities of the graph operators conditioning on their local neighboring sketch graph structures. In experiments, we learn the parameters and decision thresholds through human experiments, and we show that the sketch pyramid is a more parsimonious representation than a multi-resolution Gaussian/Wavelet pyramid. We also demonstrate an application on adaptive image display—showing a large image in a small screen (say PDA) through a selective tour of its image pyramid. In this application, the sketch pyramid provides a means for calculating information gain in zooming-in different areas of an image by counting a number of operators expanding the primal sketches, such that the maximum information is displayed in a given number of frames. A short version was published in ICCV05 (Wang et al. 2005).     

TDFA:一种生成空间影像金字塔的方法

目的 降采样滤波是生成空间金字塔影像数据的主要手段,但目前没有一种客观指标来鉴别滤波器的降采样效果,因为至少需要空间金字塔的两层原始信号才能计算滤波器的降采样峰值信噪比(PSNR)。为解决此难题,本文建立一种研究路线:先基于视频影像数据评选确定一个性能优秀的降采样滤波器,然后验证该滤波器降采样生成遥感金字塔的主观目视效果,提出一种沿图像纹理方向滤波的降采样方法TDFA(texture direction filtering approach),可生成高质量的空间影像金字塔。方法 本文把降采样与升采样结合提出一种重采样滤波对偶RSFP(re-sampling filter pair),作为当前层金字塔数据的一个逼近,用来评价降采样滤波器效果。基于RSFP评价手段,筛选出一种基于纹理滤波的金字塔生成方法TDFA:对每个8×8块,TDFA在直流、水平、135°、垂直和45°等5个方向中搜索确定图像的一个纹理方向,用一个3阶滤波器沿纹理方向实施降采样,效果优于目前最好的最邻近插值方法,无任何伪彩、锯齿、块效应或马赛克。结果 利用大量影像数据实验,同几个典型滤波器的降采样效果对比,TDFA提升平均PSNR的范围,对拉格朗日滤波器是7.29~8.44 dB;对双线性滤波器是6.26~7.40 dB;对AVS的1/4插值滤波器是5.80~6.84 dB;对最邻近插值是4.51~5.70 dB。结论 本文提出的纹理滤波降采样算法可以生成质量优于现有最好水平的遥感金字塔影像,也可以生成高质量的多层视频流媒体数据。所提出的重采样滤波对偶RSFP可以输出当前层的高精度预测,用于可伸缩视频编码处理。     

The Grid-Pyramid: A Generalized Pyramid Network

The Pyramid network is a desirable network topology used as both software data-structure and hardware architecture. In this paper, we propose a general definition for a class of pyramid networks that are based on grid connections between the nodes in each level. Contrary to the conventional pyramid network in which the nodes in each level form a mesh, the connections between these nodes may also be according to other grid-based topologies such as the torus, hypermesh or WK-recursive. Such pyramid networks form a wide class of interconnection networks that possess rich topological properties. We study a number of important properties of these topologies for general-purpose parallel processing applications. In particular, we prove that such pyramids are Hamiltonian-connected, i.e. for any arbitrary pair of nodes in the network there exists at least one Hamiltonian path between the two given nodes, and pancyclic, i.e. any cycle of length 3, 4 … and N, can be embedded in a given N-node pyramid network. It is also proven that two link-disjoint Hamiltonian cycles exist in the torus-pyramid and hypermesh-pyramid networks.     

特征注意金字塔调制网络的视频目标分割

目的 视频目标分割是在给定第1帧标注对象掩模条件下,实现对整个视频序列中感兴趣目标的分割。但是由于分割对象尺度的多样性,现有的视频目标分割算法缺乏有效的策略来融合不同尺度的特征信息。因此,本文提出一种特征注意金字塔调制网络模块用于视频目标分割。方法 首先利用视觉调制器网络和空间调制器网络学习分割对象的视觉和空间信息,并以此为先验引导分割模型适应特定对象的外观。然后通过特征注意金字塔模块挖掘全局上下文信息,解决分割对象多尺度的问题。结果 实验表明,在DAVIS 2016数据集上,本文方法在不使用在线微调的情况下,与使用在线微调的最先进方法相比,表现出更具竞争力的结果,J-mean指标达到了78.7%。在使用在线微调后,本文方法的性能在DAVIS 2017数据集上实现了最好的结果,J-mean指标达到了68.8%。结论 特征注意金字塔调制网络的视频目标分割算法在对感兴趣对象分割的同时,针对不同尺度的对象掩模能有效结合上下文信息,减少细节信息的丢失,实现高质量视频对象分割。     

Simultaneous speckle reduction and contrast enhancement for ultrasound images: Wavelet versus Laplacian pyramid

Speckle can be described as random multiplicative noise. It hampers the perception and extraction of fine details in the image. Speckle reduction techniques are applied to ultrasound images in order to reduce the noise level and improve the visual quality for better diagnoses. It is also used as preliminary treatment before segmentation and classification. Several methods have been proposed for speckle reduction in ultrasound images. Multiscale contrast enhancement has proven to be very efficient for x-ray images. A recent study by Dippel et al. doing a comparison, contrast enhancement of radiographs (x-ray and mammography), between the Laplacian pyramid and the wavelet one proves that the Laplacian pyramid method gives a better result than the wavelet one; the filtering aspect was not taken into account. In ultrasound images a strong contrast variation exists which is different from x-ray and mammography. In this paper a wavelet pyramid with simultaneous speckle reduction and contrast enhancement was applied for the first time on ultrasound images with the area of interest and compared to a Laplacian enhancement pyramid. The optimum choice of wavelet bases for ultrasound images is investigated in this study. In order to realize a fair comparison, the same nonlinear modification in both multiscale schemes is used. The comparison proves that the wavelet pyramid gives a much better result than the Laplacian one for simultaneous speckle reduction and contrast enhancement of ultrasound images. The text was submitted by the author in English. Ali Samir Saad, 1964. 1996 PhD in image processing, Polytechnics School of the Engineering University of Nantes, France. 1993 Masters in Electronics. 1990 Masters in Digital Image Processing, Institute of Computer Sciences and Communication University of Rennes, France. 1989 BS in Electrical Engineering, University of Saint-Etienne. Academy of Lyon, France. 1996–2000 Research associate at the National Center for Macromolecular Imaging. Baylor, Houston, Texas. Assistant professor at King Saud University, Dept. of Biomedical Technology. Area of research in medical image processing and analysis, 23 publications, member of the American Association for the Advancement of Sciences. Marquis Who’s Who in the World; Cambridge Blue Book 2006.     

Extension of the Nicholls-Lee-Nichols algorithm to three dimensions

A new algorithm for clipping a line segment against a pyramid in E ³ is presented. This algorithm avoids computation of intersection points that are not end points of the output line segment. It also solves all cases more effectively. The performance of this algorithm is shown to be consistently better than that of existing algorithms, including the Cohen–Sutherland, Liang–Barsky, and Cyrus–Beck algorithms.     

Interflation of functions of three variables on pyramids with one curvilinear face

Operators of interflation of a function on a pyramid with one curvilinear face are proposed for the first time. Moreover, a relationship for the approximation error and also an estimate of this error are found for the case when the curvilinear face degenerates into a plane. __________ Translated from Kibernetika i Sistemnyi Analiz, No. 6, pp. 32–49, November–December 2005.     

A Robust and Fast Non-Local Means Algorithm for Image Denoising

In the paper,we propose a robust and fast image denoising method.The approach integrates both Non- Local means algorithm and Laplacian Pyramid.Given an image to be denoised,we first decompose it into Laplacian pyramid.Exploiting the redundancy property of Laplacian pyramid,we then perform non-local means on every level image of Laplacian pyramid.Essentially,we use the similarity of image features in Laplacian pyramid to act as weight to denoise image.Since the features extracted in Laplacian pyramid are localized in spatial position and scale,they are much more able to describe image,and computing the similarity between them is more reasonable and more robust.Also,based on the efficient Summed Square Image(SSI)scheme and Fast Fourier Transform(FFT),we present an accelerating algorithm to break the bottleneck of non-local means algorithm—similarity computation of compare windows.After speedup,our algorithm is fifty times faster than original non-local means algorithm.Experiments demonstrated the effectiveness of our algorithm.     

机载SAR图像的动态金字塔实时显示技术

目的 随着机载SAR（合成孔径雷达）图像分辨率越来越高,幅宽越来越大,传统雷达显控系统将整幅图像放入内存、抽样显示的现有方法存在内存资源紧张、显示图像的等待时间过长等问题,为解决此类问题,提出一种动态金字塔实时显示技术。方法 机载SAR图像实时显示软件包括动态金字塔构建和显示技术。动态金字塔构建技术包括：当接收到一个瓦片的图像数据时,输出第0层级的金字塔瓦片;分6种情况,生成高层级瓦片,随着接收数据不断增多,逐步补全金字塔文件。动态金字塔显示技术是指在瓦片数据不全的情况下,采用递归算法,读取较低层级瓦片,合成、显示当前显示层级图像的技术。这两种技术分属两个独立线程,以硬盘文件（瓦片）为接口,实时交互,协同工作。结果 机载SAR图像实时显示软件仅仅占用30 MB内存,且与图像大小无关;显示第1块SAR图像瓦片的时延小于1 s,与传统显控系统对比,减少约一帧图像的传输时延;显示整帧图像的时延因存储介质读写文件的速率存在差异较大,固态硬盘的时延比较稳定,显示1 GB图像的时延为12.55 s;机械盘的时延受读写速度的影响,当发送时间间隔大于6 ms时,显示1 GB图像的时延仅比传输时延多1.47 s。结论 机载SAR图像实时显示软件能实时向用户呈现接收中的SAR图像,提高了机载SAR图像的显示时效性,降低了机载雷达显控终端的内存需求,改善了机载雷达显控终端的用户体验。     

不产生精度截断及数据膨胀的图像金字塔

图像金字塔是处理和分析数字图像的重要工具。以图像金字塔为研究对象,旨在减小图像金字塔的存储量,同时解决恢复数据时的精度截断问题。从同一思想出发,提出了两种新的金字塔结构：类均值金字塔与类高斯金字塔;它们都是通过对子像素的简单运算来求得父像素,而这种运算不存在精度截断问题,也不需要额外的比特来保存小数。由于从父像素能够反推出它的子像素之一,所以,在存储金字塔时可以舍弃这部分子像素,从而使所需存储的像素数仅与原始图像相同。对精度、存储量以及构建速度的理论分析与实验结果表明,该金字塔结构具有优良的总体性能。     

Design and fabrication of differently shaped pyramids on Si{100} by anisotropic wet etching

We aimed to produce differently shaped pyramids, that is, eight-sided, triangular, and rhombic pyramids, on the same Si{100} wafer by simply changing mask patterns. A triangular pyramid has an advantage in that it can always become sharp because its vertex becomes a point and is not affected by fabrication errors. A rhombic pyramid that looks like an arrow head was designed to decrease the insertion friction and scar area even if it is inserted deeply into the skin. Triangular and H-shaped etching mask patterns were designed to produce triangular and rhombic pyramid structures on Si{100} on the basis of eight-sided pyramid formation. Both triangular and rhombic pyramids were successfully fabricated as new MEMS structures by applying TMAH anisotropic wet etching (25.0 wt%, 70 °C). A sharp tip with a radius of less than a few hundred nm was obtained in both the triangular and rhombic Si pyramids.     

Bottom-up pyramid cellular acceptors with four-dimensional layers

In 1967, M. Blum and C. Hewitt first proposed two-dimensional automata as a computational model of two-dimensional pattern processing. Since then, many researchers in this field have been investigating the many properties of two- or three-dimensional automata. In 1977, C.R. Dyer and A. Rosenfeld introduced an acceptor on a two-dimensional pattern (or tape) called the pyramid cellular acceptor, and demonstrated that many useful recognition tasks are executed by pyramid cellular acceptors in a time which is proportional to the logarithm of the diameter of the input. They also introduced a bottom-up pyramid cellular acceptor, which is a restricted version of the pyramid cellular acceptor, and proposed some interesting open problems about bottom-up pyramid cellular acceptors. On the other hand, we think that the study of four-dimensional automata has been meaningful as the computational model of four-dimensional information processing such as computer animation, moving picture processing, and so forth. In this article, we investigate bottom-up pyramid cellular acceptors with four-dimensional layers, and show some of their accepting powers.     

HEXHOOP: Modular Templates for Converting a Hex-Dominant Mesh to an ALL-Hex Mesh

This paper presents a new mesh conversion template called HEXHOOP, which fully automates a con-version from a hex-dominant mesh to an all-hex mesh. A HEXHOOP template subdivides a hex/prism/pyramid element into a set of smaller hex elements while main-taining the topological conformity with neighboring elements. A HEXHOOP template is constructed by assembling sub-templates, cores and caps. A dicing template for a hex and a prism is constructed by choosing the appropriate combination of a core and caps. A template that dices a pyramid without losing conformity to the adjacent element is derived from a HEXHOOP template. Some experimental results show that the HEXHOOP templates successfully convert a hex-dominant mesh into an all-hex mesh. ID="A1" Correspondence and offprint requests to: K. Shimada, Department of Mechanical Engineering, Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh, PA 15213–3890, USA. E-mail: shimada@cmu.edu     

Two-node-Hamiltonicity of enhanced pyramid networks

An enhanced pyramid network is an alternate hierarchical structure for a pyramid network. This structure is created in a pyramid network by replacing each mesh with a torus at layers greater than one. This work studies the fault-tolerant Hamiltonian problem on the enhanced pyramid network and demonstrates that an enhanced pyramid network with two faulty nodes is Hamiltonian. The result is optimal, because edge connectivity and node connectivity of the enhanced pyramid network are both 4.     

Top-down cellular pyramids

A cellular pyramid is an exponentially tapering stack of arrays of processors (‘cells’), where each cell is connected to its neighbors (‘siblings’) on its own level, to a ‘parent’ on th level below. It is shown that in some situations, if information flows top-down only, from fathers to sons, then a cellular pyramid may be no faster than a one-level cellular array; but it may be possible to use simpler cells in the pyramid case.     

Degree-Constrained Pyramid Spanners

A t-spanner of a pyramid network is a subnetwork in which every two nodes that were connected by an edge in the original pyramid can be connected by a path in the subnetwork with at most t edges. We give several results that present trade-offs between t and the maximum degree of a t-spanner.     

Boundary localization in an image pyramid

A pyramid is a stack of successively smaller versions of a given image, with size and resolution decreasing exponentially. Large objects can be detected on high levels of the pyramid at low cost, since the images at high levels are small, but the boundaries of objects detected in this way are not accurately localized. In principle, we can “project” a boundary down to the next lower level of the pyramid; use local search at that level to localize it more precisely; and repeat the process at successively lower levels. In practice, however, it is not clear what cost function should be used in defining the “optimal” boundary at each level. We show that with proper choice of criterion, the boundaries obtained in the full-resolution image (at the base of the pyramid) seem very reasonable perceptually.     

基于拉普拉高斯金字塔的数字图像融合研究

阐述了基于拉普拉斯金字塔图像融合的原理和方法,首先对原图像分别进行拉普拉斯金字塔分解,接着在对应的分层子图像上采用变换区域特征量测法进行融合,然后对得到的拉普拉斯金字塔反变换得到融合图像。通过使用MATLAB软件对本文研究的融合算法进行仿真实验,实验结果证明基于拉普拉斯金字塔图像融合方法得到的融合图像视觉效果良好,可视化清晰,图像质量得到了较大提高,该技术可广泛应用于多种研究领域。