Showing papers in "Computer Graphics and Image Processing in 1976"
••
TL;DR: This work presents a survey of processing techniques for presenting continuous tone still images on such displays by dividing an image into picture elements and comparing the intensity of each element with a threshold value.
633 citations
••
TL;DR: Region growing systems and their role in pictorial segmentation are reviewed; developmental trends relating these systems are examined and a global initialization technique is suggested.
610 citations
••
TL;DR: A texture operator for use in computer vision programs that classifies texture according to characteristics of the Fourier transform of local image windows, using gradients of the texture as a depth cue for longitudinal surfaces.
301 citations
••
TL;DR: A regular decomposition of picture area into successively smaller quadrants is defined, which involves logarithmic search, and the resulting picture representation is shown to enable rapid access of image data without regard to position, efficient storage, and approximate structural descriptions of constituent patterns.
290 citations
••
TL;DR: Use of a number of progressive, closely spaced views has the advantage of reliability over the use of just two views for stereo, and potential savings of computational effort in analysis of scenes of three-dimensional objects.
151 citations
••
TL;DR: Through this transformational stage, the statistical approach to texture is shown to be a generalization of the structural approach in the same way that random variables are ageneralization of deterministic variables.
141 citations
••
TL;DR: In this paper, an algorithm is presented for reordering and interpolating a set of fan beam projections into an equivalent set of equispaced parallel ray projections, which may then be reconstructed by the convolution method of Bracewell and Riddle as implemented by Ramachandran and Lakshminarayanan.
73 citations
••
TL;DR: In this article, alternative iterative algorithms (conjugate gradient, nonstationary Richardson, semi-iterative) for quadratic optimization and compare their performance in image reconstruction with the previously used methods.
73 citations
••
TL;DR: In this paper, simple methods of generating digital approximations to circles and other types of curves are described, and simple methods for generating digital approximation to curves and other kinds of curves.
71 citations
••
IBM1
TL;DR: A class of black-white image compression algorithms is described which is based on chain-link encoding of the boundaries of objects, able to operate on completely arbitrary images, as opposed to special classes (e.g., line drawings).
49 citations
••
TL;DR: The distortion of pictorial features brought about by one such averaging transformation is investigated and the smearing of spots and edges, object separability, and distortion of features detectable with templates are discussed.
••
TL;DR: A context sensitive language over {0, 1} + is described whose words represent straight lines in the plane whose generation and recognition algorithms are of interest for pattern recognition, computer graphics, parallel computation, and neural net theory.
••
TL;DR: In this article, a grammar for generating hexagonal arrays on a triangular grid is presented, and a new kind of catenation, called arrowhead catenations, is defined so that the hexagonal shape is maintained in every generation.
••
TL;DR: In this paper, a new edge detector which is better than the gradient-type edge detectors used so far is presented, where the edge detector is appended to an image, a picture is obtained consisting of edge points.
••
TL;DR: In this article, a parallel iterative algorithm was devised to process the edge vectors derived from the gradient, which can be used for edge detection in digital pictures, and the resulting output consists of thinned edges which may be tracked very quickly.
••
TL;DR: Topics covered include transforms and filtering; compression; enhancement, restoration, and reconstruction; hardware and software implementations; pictorial pattern recognition; matching and local feature detection; segmentation; shape and texture; scene analysis; and formal models.
••
TL;DR: A region growing process is described, suitable for an initial, syntactic overview of a visual scene, designed to function on scenes where nonuniform surfaces discourage absolute thresholding techniques.
••
TL;DR: A shrinking algorithm for parallel processing of binary patterns that shrinks any pattern symmetrically from all sides to an isolated point at the center irrespective of the complexity of the pattern.
••
IBM1
TL;DR: Experimental digital methods for processing continuous tone pictures to produce outputs which are suitable for conventional printing methods are described and the results are shown.
••
TL;DR: A collection of simple procedures based on masks or templates to be used in computer picture processing is presented and examples of applications to various problems associated with picture processing are given.
••
TL;DR: In this article, the reconstruction of a two-dimensional image from several one-dimensional projections is simulated on a computer using filtered back projection as an algorithm, and the reconstruction error is tested under the assumption of several conditions.
••
TL;DR: A new type of plotter for computer graphics in color that uses three electrically modulated ink jets to generate both line graphics and pictorial images on ordinary paper at high speed is described.
••
••
TL;DR: A system is built that accepts pictures defined by vertices and the connections between them and transforms the pictures to digital ones and includes an algorithm that finds the simple domains or the “regions” of a picture.
••
TL;DR: The paper describes how “optimum” deconvolutional processing arrays can be efficiently computed and quantitative results are given for Airy and Gaussian spot compensation under a variety of image sampling raster dimensions and patterns.
••
TL;DR: Methods of introducting local tension into a bicubic surface patch are discussed and the ability to reduce are length along boundary curves allows the user to change the surface area of a patch.
••
TL;DR: In this paper, a statistical approach, taking noise into account, in which the image signal is taken to be a Gaussian Markovian process, is presented, and results of constant speed smear detection are also given.
••
TL;DR: Two classification schemes that use prototype pictures are discussed, along with a method for obtaining and evaluating these prototypes.
••
TL;DR: How pattern recognition can be simplified greatly by the introduction of the notion of command operators, which indicate, at one level of recognition, what operators are to be used at the preceding level, and then keep only the information needed for the recognition.
••
TL;DR: This method of recording enables the computer user to add sound to his movies using equipment and techniques with which he is familiar, and introduces problems from two sources, the incremental movement of the film and the detailed control over the recorded wave shapes.