Correspondence Problem

In structural pattern recognition, it is usual to compare a pair of objects through the generation of a correspondence between the elements of each of their local parts. To do so, one of the most natural ways to represent these objects is through attributed graphs. Several existing graph extraction methods could be implemented and thus, numerous graphs, which may not only differ in their nodes and edge structure but also in their attribute domains, could be created from the same object. Afterwards, a matching process is implemented to generate the correspondence between two attributed graphs, and depending on the selected graph matching method, a unique correspondence is generated from a given pair of attributed graphs. The combination of these factors leads to the possibility of a large quantity of correspondences between the two original objects. This paper presents a method that tackles this problem by considering multiple correspondences to conform a single one called a consensus correspondence, eliminating both the incongruences introduced by the graph extraction and the graph matching processes. Additionally, through the application of an online learning algorithm, it is possible to deduce some weights that influence on the generation of the consensus correspondence. This means that the algorithm automatically learns the quality of both the attribute domain and the correspondence for every initial correspondence proposal to be considered in the consensus, and defines a set of weights based on this quality. It is shown that the method automatically tends to assign larger values to high quality initial proposals, and therefore is capable to deduce better consensus correspondences.

In this paper, the three-dimensional reconstruction of consecutive brachial plexus slices is discussed. To get precise results, the contour-based reconstruction method is adopted. As the contour correspondence problem is of great importance to the reconstruction, a cluster method is firstly used to get rid of the offsets between the contours introduced during the stage of data acquisition. Besides correspondence check, a robust similarity-based correspondence algorithm is proposed which results in an automatic correspondence rate more than 97%.

VIRIM, a real-time direct volume rendering system is presented. The system is freely programmable and supports models like α-compositing, front-to-back (back-to-front) techniques, and the slab method. The hardware system is divided into two units, a geometry unit and a raycast unit. The geometry unit performs resampling and gradient estimation and is mapped directly into hardware. It supports different resampling filters in order to reduce resampling artifacts. The raycast unit consists of 16 digital signal processors that perform the programmable ray-casting. The software of VIRIM is layered and provides manipulation tools for the data during realtime visualization like arbitrary gray-value mapping and setting the region of interest. The system is under test and will be available as prototype 1995.

The functional equation relating the eigenvalues of the block Symmetric Successive Overrelaxation (SSOR) iteration matrix with those of the block Jacobi iteration matrix found by Chong and Cai (1985) is used in order to obtain precise domains of convergence of the block SSOR iteration method associated with a class of generalized consistently ordered (GCO) ( k , p − k )-matrices A ( p ⩾2, k =1, …, p −1). We show that the domain of convergence depends only on the relaxation parameter ω, the spectral radius of the block Jacobi iteration matrix and the value of l = k p . Unlike the case ( k , p − k ) = (1, p −1), p ⩾3, which was studied in an earlier paper by the authors (1989), beyond certain critical values of l the domain of convergence does not grow monotonically with l and manifests various sorts of behavior . However, as in the case ( k , p − k ) = (1, p −1), it is shown that the intersection of the convergence domains taken over all pairs ( p , k ) coincides with the exact domai...

Suppose that A ∈ C n,n is a block p -cyclic consistently ordered matrix and let B and S ω denote the block Jacobi and the block symmetric successive overrelaxation (SSOR) iteration matrices associated with A , respectively. Extending previous work by Hadjidimos and Neumann, the present authors have determined the exact regions of convergence of the SSOR method in the ( ϱ ( B ), ω )-plane, for any p ⩾ 3, under the further assumption that the eigenvalues of B p are real of the same sign. In the present work the investigation goes on further, several questions are raised and among others the problem of the determination of the optimal regions of convergence in the spirit of Niethammer and Varga as well as that of the optimal relaxation factor are examined.

Previous studies have indicated that the formation of coherent patterns for multielement motion displays depends onglobal cooperative interactions among large ensembles of spatially distributed motion detectors. These interactions enhance certain motion directions and suppress others. It is reported here that perceiving one element moving between two nearby locations likewise is subject to cooperative influences (possibly facilitating and inhibiting interactions within alocal ensemble of overlapping detectors). Thresholds depending on luminance contrast were measured for a generalized singleelement apparent-motion stimulus, and evidence for spontaneous switching and hysteresis effects indicated that motion perception near the 50% threshold was bistable. That is, for conditions in which motion and nonmotion were perceived half the time, the two percepts were distinct; when one was perceived, it clearly was discriminable from the other. These results indicated that (1) single-element ...

Indirect SLAM (Simultaneous Localization and
Mapping) algorithms primarily rely upon image features to
estimate poses and to perform loop closure. Traditional image
key point extraction algorithms fail to detect features in the
image in some cases due to various reasons like low exposure,
lack of contours and hardware constraints. To tackle this issue,
we propose a U-Net neural network based image keypoint
extraction algorithm that works on low cost camera hardware.
We utilize an unconventional approach by applying image
segmentation techniques to extract Shi-Tomashi features for
training the neural network. It is an attempt to building
innovative approaches for Deep learning-based visual odometry
problem. The primary focus of this work is to establish a new
research direction for deep learning based keypoint extraction
tailored for real-time SLAM on low-cost and low-power devices.
We aim to lay the foundations for further investigation into
segmentation-based key-point modeling and its impact on
critical SLAM components like feature tracking, mapping, and
loop closure detection.

Credal networks generalize Bayesian networks relaxing numerical parameters. This considerably expands expressivity, but makes belief updating a hard task even on polytrees. Nevertheless, if all the variables are binary, polytree-shaped credal networks can be efficiently updated by the 2U algorithm. In this paper we present a binarization algorithm, that makes it possible to approximate an updating problem in a credal net by a corresponding problem in a credal net over binary variables. The procedure leads to outer bounds for the original problem. The binarized nets are in general multiply connected, but can be updated by the loopy variant of 2U. The quality of the overall approximation is investigated by promising numerical experiments.

In this presentation a new search method is proposed to improve the speed and accuracy of surface based registration algorithms. Furthermore, a parallel point projection based, multicomponent distance evaluation method is presented. This method offers an elegant solution to the proble:m of partially overlapping data sets. An adaptive outlier treatment method is also presented. Combination of all these new techniques results in a faster surface based algorithm with better accuracy, but above all with better rehabihty than existing surface based 3D registration algorithms In the context of the surface correspondence problem, surface based registration algorithms are compared to feature matching methods.

We have developed a motion analysis method that combines the wavelet transform with the ¯ow surface image technique. In dynamic navigation environments, whenever objects move, the projections of the objects onto the image plane also move. These projections build over time spatiotemporal surfaces of their movement and volumes created by the surfaces. This paper presents a new method for the interpretation of optical ¯ow for moving objects from a sequence of images. Flow surface images of the moving objects are created within the wavelet-derived space, chosen from seven dierent directionally sensitive detail images using the 3D wavelet decomposition. The motion estimation algorithm concentrates on the integration of information from the ¯ow surface images, followed by a quadratic patch parameterization and determination of ¯ow paths of the end points of edges on the ¯ow surface images. The results of two experimental studies with an object exhibiting out-of-plane translation and rotation are also reported.

Molecular dynamics (MD) simulations are employed to investigate the capillary fluctuations of steps on the surface of a model metal system. The fluctuation spectrum, characterized by the wave number (k) dependence of the mean squared capillary-wave amplitudes and associated relaxation times, is calculated for 〈110〉 and 〈112〉 steps on the {111} surface of elemental copper near the melting temperature of the classical potential model considered. Step stiffnesses are derived from the MD results, yielding values from the largest system sizes of (37±1)meV/A[over ˚] for the different line orientations, implying that the stiffness is isotropic within the statistical precision of the calculations. The fluctuation lifetimes are found to vary by approximately four orders of magnitude over the range of wave numbers investigated, displaying a k dependence consistent with kinetics governed by step-edge mediated diffusion. The values for step stiffness derived from these simulations are compared ...

Learning robust feature descriptors from multiple views plays an important role in image matching and its downstream tasks. Due to various uncertainty factors from illumination to viewpoint, robust feature matching remains a challenging task; the generalization property of learned feature descriptors has also remained poorly understood. Our intuition is that ensemble learning can assemble a collection of feature descriptors by mining richer information for each pixel, which can lead to improved robustness and generalization properties. In this paper, we propose a new image feature description method named Orthogonal Descriptor Network (OD-Net), which describes a pixel by a multi-branch structure and then fuses them. To avoid the model being trapped into an ill-posed solution and encourage the network to mine complementary information, we have designed an orthogonality constraint and modeled it as a novel loss function. It is worth noting that the idea of orthogonal feature extraction is general and can be easily plugged into many existing frameworks. Extensive experiments have been carried out to show that our OD-Net can produce better results than the current state-of-the-art on a variety of image-matching tasks and evaluation metrics.

This paper presents a new approach to bitext correspondence problem (BCP) of noisy bilingual corpora based on image processing (IP) techniques. By using one of several ways of estimating the lexical translation probability (LTP) between pairs of source and target words, we can turn a bitext into a discrete gray-level image. We contend that the BCP, when seen in this light, bears a striking resemblance to the line detection problem in IP. Therefore, BCPs, including sentence and word alignment, can benefit from a wealth of effective, well established IP techniques, including convolution-based filters, texture analysis and Hough transform. This paper describes a new program, PlotAlign that produces a word-level bitext map for noisy or non-literal bitext, based on these techniques.

This paper develops a novel solution to the problem of recovering the structure of a scene given an uncalibrated video sequence depicting the scene. The essence of the technique lies in a method for recovering the rigid transformation between the different views in the image sequence. Knowledge of this 3D motion allows for self-calibration and for subsequent recovery of 3D structure. The introduced method breaks away from applying only the traditionally used epipolar constraint and introduces a new constraint based on the interaction between 3D motion and shape. Up to now, structure from motion algorithms proceeded in two well defined steps, where the first and most important step is recovering the rigid transformation between two views, and the subsequent step is using this transformation to compute the structure of the scene in view. Here both aforementioned steps are accomplished in a synergistic manner. Existing approaches to 3D motion estimation are mostly based on the use of optic flow which however poses a problem at the locations of depth discontinuities. If we knew where depth discontinuities were, we could (using a multitude of approaches based on smoothness constraints) estimate accurately flow values for image patches corresponding to smooth scene patches; but to know the discontinuities requires solving the structure from motion problem first. In the past this dilemma has been addressed by improving the estimation of flow through sophisticated optimization techniques, whose performance often depends on the scene in view. In this paper the main idea is based on the interaction between 3D motion and shape which allows us to estimate the 3D motion while at the same time segmenting the scene. If we use a wrong 3D motion estimate to compute depth, then we obtain a distorted version of the depth function. The distortion, however, is such that the worse the motion estimate, the more likely we are to obtain depth estimates that are locally unsmooth, i.e., they vary more than the correct ones. Since local variability of depth is due either to the existence of a discontinuity or to a wrong 3D motion estimate, being able to differentiate between these two cases provides the correct motion, which yields the "smoothest" estimated depth as well as the image locations of scene discontinuities. Although no optic flow values are computed, we show that our algorithm is very much related to minimizing the epipolar constraint when the scene in view is smooth.