Showing papers on "Orientation (computer vision) published in 1977"

Display of three-dimensional information in computed tomography.

Abstract: From a sequence of computed tomograms of two-dimensional transverse slices of a body one can build up a three-dimensional array of numbers containing spatial rather than cross-sectional information. We have developed computer methods to display a slice at arbitrary orientation through the three-dimensional object. The three-dimensional array of numbers can also be used to display what any particular organ would look like if it were removed from the body. This is done by a computer technique that first detects the surface of the organ of interest and then displays this surface on a screen.

Representing Visual Information

Abstract: : Vision is the construction of efficient symbolic descriptions from images of the world An important aspect of vision is the choice of representations for the different kinds of information in a visual scene In the early stages of the analysis of an image, the representations used depend more on what it is possible to compute from an image than on what is ultimately desirable, but later representations can be more sensitive to the specific needs of recognition This essay surveys recent work in vision at MIT from a perspective in which the representational problems assume a primary importance An overall framework is suggested for visual information processing, in which the analysis proceeds through three representations; (1) the primal sketch, which makes explicit the intensity changes and local two-dimensional geometry of an image, (2) the 2 1/2-D sketch, which is a viewer-centered representation of the depth, orientation and discontinuities of the visible surfaces, and (3) the 3-D model representation, which allows an object-centered description of the three-dimensional structure and organization of a viewed shape Recent results concerning processes for constructing and maintaining these representations are summarized and discussed (Author)

Autonomous, check-pointing, navigational system for an airborne vehicle

Abstract: An autonomous navigational system for an airborne vehicle is disclosed which automatically updates its position and heading coordinates at each of a plurality of checkpoints along a preselected flight path. Prestored in a memory of the navigational system is a plurality of reference map images respectively associated with the plurality of checkpoints. Each reference map image is comprised of a plurality of terrain features associated with the terrain about an associated checkpoint. In operation, an acquisition unit senses an image of local terrain at each checkpoint. The relative sizes of reference and sensed map images vary with the mission requirements. A terrain feature extractor detects a plurality of terrain features contained in the sensed map image. The detected terrain features in the sensed map image are selectively matched in a feature pattern matcher against the terrain features of the associated reference map image to determine the position and orientation of the sensed map image within the associated reference map image. The output of the feature pattern matcher comprises error signals which are used by an inertial guidance system to update the position and heading coordinates of the airborne vehicle.

The contouring of orientation data represented in spherical projection

Abstract: The inadequacy of the current method of contouring orientation data in terms of the concentration of points occurring in 1% of the projected area is demonstrated. In particular, the maximum concentration and the areas occupied by the different concentrations are shown to be dependent on sample size. Using a counting area of 100/n% of the projection, where n is the sample size, yields meaningful orientation patterns which are comparable for different sample sizes. This is illustrated with reference to computer generated random and non-random orientation patterns. The data obtained are compared with predicted values and form a basis for comparison with measured orientation patterns to test their significance.

Model-based vision system for scenes containing multiple parts

Abstract: A model-based computer vision system has been developed which can determine the position and orientation of several parts in the same scene. The parts can be of the same type or of different types. Furthermore, the system can work with visually noisy scenes and some occlusion. Abstract two-dimensional models are formed by having the system view the parts under favorable conditions. Parts are found by matching the abstract models to the scene data, after it has been highly organized.

Significance of operator variation and the angle of illumination in lineament analysis on synoptic images

Abstract: The significance of operator variation and the angle of illumination in acquired imagery is analyzed for lineament analysis. Five operators analyzed a LANDSAT image and four photographs of a plastic relief map illuminated at a low angle from varying directions of the Prescott, Arizona region. Significant differences were found in both number and length of the lineaments recognized by the different investigators for the images. The actual coincidence of lineaments recognized by the investigators for the same image is exceptionally low. Even the directional data on lineament orientation is significantly different from operator to operator and from image to image. Cluster analysis of the orientation data displays a clustering by operators rather than by images. It is recommended that extreme caution be taken before attempting to compare different investigators' results in lineament analysis.

Testing opaque workpieces, particularly tiny or thin elongated objects such as pins, bolts, wires, or the like, for orientation, size and material integrity

Abstract: A light source, preferably an electrical flash tube, generates a beam of light which, preferably, is collimated to provide a parallel beam output. The workpieces are placed in the beam of light to generate a shadow image. The shadow image and surrounding beam portions are projected on a screen, with intervening optical enlargement, on which photoelectric sensors are located positioned in the projected shadow image and preferably also in adjacent regions of the light beam, the outputs from the sensors being compared with each other, or with a reference, to determine relative shadow image characteristics of regions of the objects in the light beam path, the output signals being a measure of differences in characteristics, and hence determinative of orientation, size, or integrity of the workpiece material. To test for size or orientation, two sensors are located at the end portions of the shadow images of pins, or bolts, carried through a test path interposed between the parallel beams from the light source, before enlargement.

Image segmentation technique for locating automotive parts on belt conveyors

Abstract: A simple, model free computer vision program to determine the locations of non-overlapping parts on belt conveyors is described. This program illustrates a simple and effective procedure for segmenting objects from background in instances where simple thresholding of a gray-level image does not suffice. The procedure consists of a unique sequence of standard image enhancement processes. The resultant image exhibits silhouettes of the objects, which contain sufficient information for locating those parts whose orientation can be determined without observation of internal features. The technique has been implemented on a large research computer, as well as a mini-computer coupled to a prototype belt conveyor-robot arm part transfer system. The technique has been validated for a large variety of parts and belt surfaces. It can meet production rates and has the potential for actual production use.

Method of orienting objects using optically smeared images

Abstract: The orientation of an industrial part such as a microcircuit chip having a pattern of dominant parallel lines is determined by an optical method. An image of the parallel line pattern is smeared or defocused in a known direction such that the resultant smeared image of the part reveals the lines in sharp detail when they are parallel to the known direction of smearing and are blurred if the lines are angularly displaced from the known direction. By rotating the image of the pattern relative to the smear direction and determining the angle at which maximum sharpness occurs, the angular orientation is determined. By correlating the pattern of sharp lines of the smeared image with a similar reference pattern of known position, the positional orientation of the part is also determined. When two angularly displaced families of parallel lines are present on the object, positional coordinates relative to two axes are determined.

Method and means for locating process points on miniaturized circuits

Abstract: A method and an apparatus is provided by which automatic or semi-automatic processing of integrated and printed circuit devices can be accomplished notwithstanding that the points at which processing must be accomplished have been displaced from their assigned positions by translation and rotation The method involves measurement of displacement of some circuit points from their designated positions in a specified sequence That is coupled with a physical positioning and orientation of the device relative to the process apparatus and with approximate calculation An apparatus is disclosed by which the measurements, sequence of measurements, orientation, and computation required can be accomplished

Estimation of image centroid, size, and orientation with laser radar

Abstract: Angular tracking of targets with monopulse laser radar can be performed by tracking the centroid of the image. The image irradiance distribution is sensed with an array of noncoherent photocounting detectors and the counts processed to provide estimates of the image centroid, size, and orientation. The mean and variance of the centroid and second moment are obtained for optically rough and smooth extended targets, as functions of coherence, target extent, resolution, and integration time. The limits on accuracy imposed by shot noise from the image and background irradiances and by laser speckle are discussed.

Multifacet mirror for capture of IR radiation - consists of number of individual visual fields and gives reflection onto scanning element

Abstract: The multiple facet mirror facilitates the capture of infrared radiation from a number of individual visual fields and also its reflection onto infra-red radiation scanning element. The visual fields are arranged at intervals from each other. The mirror features a number of concave surface areas which capture the radiation. Each concave surface area (37 to 46) displays an orientation for the capture and reflection of infra-red radiations from a given visual field not parallel to the axis of the concave surface. The concave surface areas (37 to 46) exhibit focal points arranged at a distance to one another and at a distance from the intended position of the scanning element.

A new method for internal calibration of left ventricular cineangiography.

Abstract: Previous methods of internal calibration for cineangiography have made use of reference objects such as catheters or wires of known length or diameter. Such devices have either insufficient X-ray resolution, or require a specific orientation within the ventricle which is difficult to confirm. External calibration methods, while eliminating these errors, can only estimate actual position of the left ventricle. The use of a Swan-Ganz balloon catheter as a calibration object diminishes these problems. When positioned within the left ventrical and filled with a radiopaque medium, the balloon is of sufficient size and density to avoid errors induced by previously employed objects. The inflated balloon geometry is simple and reproducible. The use of this internal calibration catheter markedly diminishes major sources of errors in the determination of ventriculograms for both dogs and humans.

Optical processing of structural contour maps

Abstract: Optical processing is a fast, easy, economical, and reliable method of quantifying map patterns. A reduced transparency of a zebra-like pattern, obtained by blackening in alternate contour bands of a map, is used as input to an optical system of a laser source and objective lens to produce Fourier transforms. By sampling the intensity distribution across the transform, information about the spatial frequency (periodicity) and orientation in the original map can be inferred. Parameters used for comparing the map images are the values of intensity distribution in the transform. These intensities are normalized and subjected to comparison using various similarity methods. Cross-multiplication of the intensities is suggested as a measure of vectorial comparison of the orientations. By plotting the directional intensities into a polar coordinate system, rose diagrams are prepared for visual comparison. Optically derived parameters are independent of scale and other restrictive and limiting requirements as demanded in techniques of map analyses based on numerically derived parameters. The technique can be used in analyzing other pictorial data including aerial photographs, rock-fabric diagrams, and thin sections.

Iris information setting device for camera

Abstract: PURPOSE:To provide an iris information setting device for a camera, in which a pair of comb-shaped conductors out of phase are scanned by a contact to generate pulses, thereby making it possible to determine the reliable orientation toward the iris stop by means of a simple construction.