Showing papers in "Computer Vision and Image Understanding in 1997"

TL;DR: A deformable surface algorithm is used to find a parametric representation of the outer cortical surface and then to define a map between corresponding cortical regions in two brain images, which brings two images into register based on elastically deformable models.

TL;DR: This work proposes two strategies which lead to either surfacical skeletons or wireframe skeletons, and two angular criteria are proposed that allow us to build a size-invariant hierarchy of simplified skeletons.

TL;DR: The present toolkit, the underlying statistical and geometric models, and the extensions, some already demonstrated and some currently under construction, that reach out to the additional information content of biomedical images in common clinical or scientific applications are reviewed.

TL;DR: A new technique is developed that allows identified objects in the image to move as rigid bodies, while the remainder smoothly deforms, valid in any dimension, subject to the choice of the basis function.

TL;DR: 3D–2D projective transformation (composition of a rigid displacement and a perspective projection) which maps a 3D object onto a 2D image of this object.

TL;DR: Experimental results for real medical endoscope images of the stomach wall show the feasibility of the Kimmel?Bruckstein algorithm of shape from shading to the endoscope image and its promising availability for morphological analyses of tumors on human inner organs.

TL;DR: A knowledge taxonomy is elaborated consisting of seven modules, ranging in specificity from task domain knowledge to generic knowledge about the use of software systems, and a discussion of their relative merits is presented.

TL;DR: In this article, the authors present a comprehensive discussion on connected morphological operators for binary images and show that alternating sequential filters resulting from grain openings and closings are strong filters and obey a strong absorption property.

TL;DR: In this thesis, the task of sensor planning for object search is formulated as an optimization problem and an approximate solution employing a one step look-ahead strategy is proposed, equivalent to the optimal solution under certain conditions.

TL;DR: In this article, a new edge-based approach for efficient image registration is proposed, which applies wavelet transform to extract a number of feature points as the basis for registration, each selected feature point is an edge point whose edge response is the maximum within a neighborhood.

TL;DR: A speechreading (lip-reading) system, where the extracted features are modeled by Gaussian distributions and their temporal dependencies by hidden Markov models, which achieves an accuracy about equivalent to that of untrained humans.

TL;DR: Two different methods to compute the fundamental matrix and its uncertainty at the same time are given and the definition of the epipolar band for stereo matching, projective reconstruction, and self-calibration is provided.

TL;DR: An algorithm which uses information from both surface reflectance and illumination variation to solve for color constancy, and in addition uses the variation to constrain the solution.

TL;DR: An active object recognition strategy which combines the use of an attention mechanism for focusing the search for a 3D object in a 2D image, with a viewpoint control strategy for disambiguating recovered object features is presented.

TL;DR: A technique for shape representation and the recognition of objects based on multiscale curvature information that provides a single framework for both the decomposition and recognition of both planar curves as well as surfaces in three-dimensional space is presented.

TL;DR: A new method based on diffusion to smooth out the noise and extract shape skeletons in a robust way and is applicable to shapes which may have junctions such as triple points.

TL;DR: An approach for 3D tag localization and tracking of SPAMM data by a novel deformable B-solid that is implemented on tag data from short axis (SA) cardiac images, as well as SA LV boundaries, and is currently being extended to include long axis data.

TL;DR: Two different sets of color-related, locally computable motion “invariants” are analyzed and tested in this paper, and the results of motion estimation based on them are compared to the direct use of the RGB brightness functions.

TL;DR: This work presents an implemented computational theory that derives dynamic descriptions of interacting objects directly from camera input, based on an analysis of the Newtonian mechanics of a simplified scene model.

TL;DR: This research derives a physics-based generalized reflectance map model which includes diffuse and specular reflection effects, a nearby point light source and perspective projection, and then develops a new direct shape recovery algorithm from shaded images which linearizes the nonlinear equations and minimizing a quadratic cost functional.

TL;DR: A new approach toward image segmentation is proposed in which a set of slightly different segmentations is derived from the same input and the final result is based on the consensus among them, using the hierarchical, RAG pyramid technique.

TL;DR: A scheme is proposed where junction candidates are first detected in a multiscale preprocessing step, which generates junction candidates with associated regions of interest that serve as input for an MDL approximation method which tests the validity of the break point hypotheses and classifies the resulting edge segments as either “straight” or “curved.”

TL;DR: The robot motion is directly driven by the first-order time-space image derivatives, which can be estimated robustly and fast and is operating in real time and the performance is robust both in theego-docking andeco-d dockingparadigms.

TL;DR: A new registration and gain correction algorithm for 3D medical images that presents the advantage of minimizing a well-defined global criterion, to deal with various classes of transformations, to be simple to implement, and to be efficient in practice is presented.

TL;DR: This paper presents an approach for integrating local evidence such as regional homogeneity and edge response to form global structure for figure?ground segmentation using a shock-based morphogenetic language.

TL;DR: The paper describes a procedure for the generation of the Blum skeleton (medial axis) of large, complex, digitized 3D objects, a 3D generalization of the Voronoi skeleton concept, which is already in routine use for 2D shapes.

TL;DR: A novel method for vectorizing line drawing images is presented, based on a sequence of a standard vectorization algorithm and maximum threshold morphology, which can be iterated until a fitting criterion is met.

TL;DR: This paper breaks down the correspondence problem into its general components: token type, match constraints, and method employed to encode and search match information and discusses at some length the benefits of hierarchical matching.

TL;DR: It is found that, in general, the optimal reconstruction of polyhedra with a bounded number of faces may take an unbounded number of silhouettes, and it is shown that O(n5) silhouettes are sufficient, and an algorithm for finding the viewpoints is described.