Showing papers in "Image and Vision Computing in 1991"

TL;DR: In this article, an approach for the detection of stationary obstacles and moving objects on the road of an autonomous vehicle by evaluation of optical flow fields from image sequences is presented, which can be interpreted to infer information about the three-dimensional environment if the camera is moving on a planar surface.

TL;DR: The Probabilistic Hough Transform, a mathematically ‘correct’ form of the Hough transform, is defined as a likelihood function in the output parameters, and a model of feature error characteristics is proposed, combining normally distributed measurement errors with uniformly distributed correspondence errors.

TL;DR: A series of tools has been developed to synthesize full colour photographic quality facial caricatures that not only provide amusement, but can be used as stimuli in psychological investigations into the representation of faces in human memory.

TL;DR: It is shown that the three-dimensional edge tracking algorithm extracts additional edges not provided by the filtering stage without introducing spurious edges.

TL;DR: An iterative scheme for improving camera calibration based on results is derived, and performance demonstrated on real data.

TL;DR: A comparative study on motion estimation from 3D line segments is presented, using both synthetic and real data obtained by a trinocular stereo, and it is observed that the extended Kalman filter with the rotation axis representation of rotation is preferable.

TL;DR: It is demonstrated that algebraic invariants of the polynomial in the fitting process can yield powerful representations of shape that are unaffected by projection, and hence make good cues for model-based vision.

TL;DR: In this paper, the authors used information internal to the stereo algorithm of Drumheller and Poggio to locate disparity discontinuities. But they did not use post-processing the stereo data.

TL;DR: The attempt is to develop correspondence-free tracking schemes and get rid of the limitations inherent in the optical flow formalism, and it is shown how a monocular observer can track an initially foveated object and keep it stationary in the centre of the visual field assuming that the target is always visible during the tracking phase.

TL;DR: Techniques for the display of natural colour images on high resolution graphics displays and adaptation of the colour quantization techniques to the case of high resolution display are described.

TL;DR: It is shown that curvatures can actually be computed this way in practice, but that they are highly sensitive to errors in viewer-motion estimates, and sensitivity is reduced, representing a significant step in the development of practical techniques for robust, qualitative 3D vision.

TL;DR: The notion that phase differences can be used to interpret disparity between a pair of stereoscopic images is applied and the derivative of phase is used to obtain the instantaneous frequency from an image, which is then used to improve the disparity estimate.

TL;DR: It is shown how a particular arrangement of fundus cameras allows for a technique akin to homomorphic filtering to recover estimates of δz δx that can be smoothed by appropriate regularization that are integrated with photometric stereo depth estimates.

TL;DR: The existence of phase singularities, and their relation to the neighbourhoods where phase is unreliable, are discussed and a simple method for detecting regions of instability is presented.

TL;DR: A system which solves the problem of relative camera geometry using an established corner detector combined with a robust stereo matching algorithm and a variational solution for the camera geometry is described.

TL;DR: This paper addresses the problems associated in processing arrays of depth data in order to achieve the goal of automatic inspection of mechanical parts, i.

TL;DR: A new class of algorithms is described for the analysis boundaries in a discrete image that is an extension of known methods for edge detection which are based on Gaussian filtering but instead of using a discrete filter and exhaustive evaluation, the filter output is computed only at isolated points.

TL;DR: A new algorithm is presented for estimating the spectral reflectance of a surface from a colour image using highlights, which is based on detailed reflectance models and is more general than existing techniques.

TL;DR: A logic program has been developed to recognise the class of a volumetric primitive object in an image description which consists of curve properties and relations, part of a scheme to recognise complex objects from a large model base.

TL;DR: A novel semi-analytic algorithm to determine the parameters of stereo camera geometry from pairs of matched points is presented that transforms the problem of locating the global minimum of a function in three parameters to one of finding the zeros of a thirteenth degree polynomial in a single parameter.

TL;DR: Structure-from-motion algorithms based on matched point-like features under orthographic projection are explored, for use in analysing image motion from small rigid moving objects, and closed-form n-point algorithms are devised that minimize image-plane positional errors.

TL;DR: A system for segmentation and grouping, based on the pseudo-Wigner distribution a discrete approximation to the WD, is discussed, and experimental results are shown.

TL;DR: A general method of generating hierarchical tilings 1 on a two-dimensional, optionally square lattice using the Tesseral Amalgamator Theory (TAT) to increase the efficiency of hierarchical methods is described.

TL;DR: The local-feature-focus method has become a standard means for robustly locating objects in two dimensions, yet it is not without its difficulties, since the maximal clique approach to graph matching which it employs is excessively computation intensive.

TL;DR: The competence of this 3D vision system is demonstrated by visually guiding a robot arm to pick up various objects in a cluttered scene with a total processing time of approximately ten seconds.

TL;DR: A sound theoretical background for feature grouping processes is offered using a Bayesian approach which makes explicit the world knowledge which is applied at any stage, and a framework which can integrate many different forms of grouping and different levels of information is described.

TL;DR: Constraints of local surface smoothness and global surface continuity are used to solve the correspondence problem, and a stereo algorithm called Needles is described, specialized to deal with smooth textured surfaces.

TL;DR: Efforts have been made to parallelize the Hough Transform, focusing on partitioning the work for a multiprocessor system, and a breakdown of the various computation overheads is presented and analyzed.

TL;DR: The design of a new image acquisition system for static scenes is presented, based on effective combination of two well known techniques of 3D vision; the triangulation method, using an image with certain areas illuminated by a controlled light plane and a striped light image lit by parallel grid projection.

TL;DR: The frame-based computer vision system named FABIUS implemented in Prolog is shown to be well suited to rapid development of applications and alternative interpretation strategies, and its operation is illustrated when applied to the task of identifying vehicles in cluttered real-world scenes.