The authors describe a general-purpose, representation-independent method for the accurate and computationally efficient registration of 3-D shapes including free-form curves and surfaces. The method handles the full six degrees of freedom and is based on the iterative closest point (ICP) algorithm, which requires only a procedure to find the closest point on a geometric entity to a given point. The ICP algorithm always converges monotonically to the nearest local minimum of a mean-square distance metric, and the rate of convergence is rapid during the first few iterations. Therefore, given an adequate set of initial rotations and translations for a particular class of objects with a certain level of 'shape complexity', one can globally minimize the mean-square distance metric over all six degrees of freedom by testing each initial registration. One important application of this method is to register sensed data from unfixtured rigid objects with an ideal geometric model, prior to shape inspection. Experimental results show the capabilities of the registration algorithm on point sets, curves, and surfaces. >

A method for registration of 3-D shapes

We describe an object detection system based on mixtures of multiscale deformable part models. Our system is able to represent highly variable object classes and achieves state-of-the-art results in the PASCAL object detection challenges. While deformable part models have become quite popular, their value had not been demonstrated on difficult benchmarks such as the PASCAL data sets. Our system relies on new methods for discriminative training with partially labeled data. We combine a margin-sensitive approach for data-mining hard negative examples with a formalism we call latent SVM. A latent SVM is a reformulation of MI--SVM in terms of latent variables. A latent SVM is semiconvex, and the training problem becomes convex once latent information is specified for the positive examples. This leads to an iterative training algorithm that alternates between fixing latent values for positive examples and optimizing the latent SVM objective function.

/pdf/object-detection-with-discriminatively-trained-part-based-4cnosvuqcp.pdf

Object Detection with Discriminatively Trained Part-Based Models

Stereo matching is one of the most active research areas in computer vision. While a large number of algorithms for stereo correspondence have been developed, relatively little work has been done on characterizing their performance. In this paper, we present a taxonomy of dense, two-frame stereo methods designed to assess the different components and design decisions made in individual stereo algorithms. Using this taxonomy, we compare existing stereo methods and present experiments evaluating the performance of many different variants. In order to establish a common software platform and a collection of data sets for easy evaluation, we have designed a stand-alone, flexible C++ implementation that enables the evaluation of individual components and that can be easily extended to include new algorithms. We have also produced several new multiframe stereo data sets with ground truth, and are making both the code and data sets available on the Web.

/pdf/a-taxonomy-and-evaluation-of-dense-two-frame-stereo-4c20etkubp.pdf

A taxonomy and evaluation of dense two-frame stereo correspondence algorithms

A comparative study of texture measures with classification based on featured distributions

Planning algorithms are impacting technical disciplines and industries around the world, including robotics, computer-aided design, manufacturing, computer graphics, aerospace applications, drug design, and protein folding. This coherent and comprehensive book unifies material from several sources, including robotics, control theory, artificial intelligence, and algorithms. The treatment is centered on robot motion planning but integrates material on planning in discrete spaces. A major part of the book is devoted to planning under uncertainty, including decision theory, Markov decision processes, and information spaces, which are the “configuration spaces” of all sensor-based planning problems. The last part of the book delves into planning under differential constraints that arise when automating the motions of virtually any mechanical system. Developed from courses taught by the author, the book is intended for students, engineers, and researchers in robotics, artificial intelligence, and control theory as well as computer graphics, algorithms, and computational biology.

Planning Algorithms: Introductory Material

In this thesis a new continuous, viewer-centered representation for polyhedral objects or scenes is introduced, called the aspect representation or asp. The asp is viewer-centered in the sense that it represents the appearance of a polyhedron to a viewer as a two-dimensional line drawing, rather than the volume of space that it fills. It is continuous in the sense that it represents appearance for all viewpoints rather than for a discrete set of viewpoints. In effect, it is a representation of appearance as a function of viewpoint. Analyses of the size of asps and algorithms for their construction are given under orthographic and perspective viewing models, for convex and non-convex polyhedra. The worst-case size of an asp is Θ(n) in the convex case and Θ(n4) in the non-convex case. The asp is the first representation of appearance as a function of viewpoint. As such it is useful for problems that depend on knowing the appearance of an object from many viewpoints, finding ranges of viewpoints with particular characteristics of appearance, and determining viewpoints at which particular visual events happen. Many problems in computer vision and graphics fall into one of these categories. In general, the asp is useful for problems that make repeated use of appearance characteristics, justifying the representation of all visual events. In this thesis the asp is applied to three problems, in each case yielding improvements or advantages over previous results. The first problem is the construction of the aspect graph, a graph defined to have a vertex for each topologically-distinct view of a polyhedron and edges connecting adjacent views. Using the asp, we present the first algorithm for constructing the aspect graph for general polyhedra. The second application is object recognition. In this application we show how to use the asp to represent and use occlusion information in recognizing three-dimensional objects from an arbitrary viewpoint. The third application is animating rotation, or showing views of a polyhedral scene from many closely-spaced viewpoints fast enough to give the appearance of the scene as the viewer moves around it. Chapter 1: Representation and Appearance 1-1. Representation for Problem Solving Any system that seeks to solve a problem must represent the problem to be solved; representation of some aspect of the world is one of the most basic requirements of a problem-solving system. Reasoning systems require a representation of knowledge and beliefs; natural language understanding systems require a representation of language and meaning; robot motion planning systems require a representation of location and shape of objects in the world. A particular representation may be more or less helpful in the solution of a problem. Choosing an advantageous representation is often a key to finding an efficient solution because an advantageous representation makes explicit the information pertinent to the solution. For example, given two equivalent knowledge databases, one may yield a proof of a theorem much more quickly than the other if it includes key lemmas. Problems in computer vision, computer graphics, CAD/CAM, robotics, and many other areas require a representation of objects in the world. Thus, it is not surprising that there are many different object representations in use with different advantages and disadvantages. Major existing three-dimensional (3-D) object representations can be divided into three main categories: volumetric, boundary, and swept-volume representations. Volumetric representations, such as octrees [Jackins & Tanimoto, 1980], space occupancy arrays [March & Steadman, 1974], [Srihari, 1981], and constructive solid geometry [Requicha, 1978], represent the volume of space that an object occupies by constructing the volume as a combination of simpler volumes. Boundary representations such as winged-edge polyhedra [Baumgart, 1972] represent the boundary of the volume. Swept-volume representations represent the cross-sectional area of the object swept along a spine [Binford, 1971]. However, all of the standard representations have something in common: they are object-centered in the sense that they represent the volume of Euclidean 3-space that the object occupies. Object-centered representations have properties that make them useful for many problems. For example, the space taken to describe an object is generally small, and it is relatively easy to find the union and intersection of two objects. However, for many applications it is important to know

The asp: a continuous, viewer-centered object representation for computer vision

Presents an approach for recovering a global surface model of an object from the deformation of the occluding contour using an active (i.e., mobile) observer able to control its motion. In particular, the authors consider two problems: (1) How can the observer's viewpoint be controlled in order to generate a dense sequence of images that allows incremental reconstruction of an unknown surface, and (2) how can one construct a global surface model from the generated image sequence? Solving these two problems is crucial for automatically constructing models of objects whose surface is non-convex and self-occludes. They achieve the first goal by purposefully and qualitatively controlling the observer's instantaneous direction of motion in order to control the motion of the visible rim over the surface. They achieve the second goal by using a calibrated trinocular camera rig and a mechanism for controlling the relative position and orientation of the viewed surface with respect to the trinocular rig. >

/pdf/building-global-object-models-by-purposive-viewpoint-control-1otn5u27kv.pdf

Building global object models by purposive viewpoint control

2D curve representations usually take algebraic forms in ways not related to visual perception. This poses great difficulties in connecting curve representation with object recognition where information computed from raw images must be manipulated in a perceptually meaningful way and compared to the representation. In this paper we show that 2D curves can be represented compactly by imposing shaping constraints in curvature space, which can be readily computed directly from input images. The inverse problem of reconstructing a 2D curve from the shaping constraints is solved by a method using curvature shaping, in which the 2D image space is used in conjunction with its curvature space to generate the curve dynamically. The solution allows curve length to be determined and used subsequently for curve modeling using polynomial basis functions. Polynomial basis functions of high orders are shown to be necessary to incorporate perceptual information commonly available at the biological visual front-end.

Perception-Based 2D Shape Modeling by Curvature Shaping

In this paper we show that Ullman and Basri's linear combination (LC) representation, which was originally proposed for alignment-based object recognition, can be used for outlier detection in motion tracking with an affine camera. For this task LC can be realized either on image frames or feature trajectories, and therefore two methods are developed which we call linear combination of frames and linear combination of trajectories. For robust estimation of the linear combination coefficients, the support vector regression (SVR) algorithm is used and compared with the RANSAC method. SVR based on quadratic programming optimization can efficiently deal with more than 50 percent outliers and delivers more consistent results than RANSAC in our experiments. The linear combination representation can use SVR in a straightforward manner while previous factorization-based or subspace separation methods cannot. Experimental results are presented using real video sequences to demonstrate the effectiveness of our LC+SVR approaches, including a quantitative comparison of SVR and RANSAC.

/pdf/linear-combination-representation-for-outlier-detection-in-kssue2eovx.pdf

Linear combination representation for outlier detection in motion tracking

This paper introduces a method for metric self-calibration that is based on a novel decomposition of the fundamental matrix between two views taken by a camera with fixed internal parameters. The method blends important advantages of the Kruppa constraints and the modulus constraint: it works directly from fundamental matrices and uses a reduced-parameter representation for stability. General properties of the new decomposition are also developed, including an intuitive interpretation of the three free parameters of internal calibration. The approach is demonstrated on both real and synthetic data.

Charles R. Dyer

Papers

The asp: a continuous, viewer-centered object representation for computer vision

Building global object models by purposive viewpoint control

Perception-Based 2D Shape Modeling by Curvature Shaping

Linear combination representation for outlier detection in motion tracking

Metric self calibration from screw-transform manifolds