Showing papers on "Orientation (computer vision) published in 1992"

A versatile camera calibration technique for high-accuracy 3D machine vision metrology using off-the-shelf TV cameras and lenses

Abstract: A new technique for three-dimensional (3D) camera calibration for machine vision metrology using off-the-shelf TV cameras and lenses is described. The two-stage technique is aimed at efficient computation of camera external position and orientation relative to object reference coordinate system as well as the effective focal length, radial lens distortion, and image scanning parameters. The two-stage technique has advantage in terms of accuracy, speed, and versatility over existing state of the art. A critical review of the state of the art is given in the beginning. A theoretical framework is established, supported by comprehensive proof in five appendixes, and may pave the way for future research on 3D robotics vision. Test results using real data are described. Both accuracy and speed are reported. The experimental results are analyzed and compared with theoretical prediction. Recent effort indicates that with slight modification, the two-stage calibration can be done in real time.

Texture Classification by Wavelet Packet Signatures.

Abstract: This correspondence introduces a new approach to characterize textures at multiple scales The performance of wavelet packet spaces are measured in terms of sensitivity and selectivity for the classification of twenty-five natural textures Both energy and entropy metrics were computed for each wavelet packet and incorporated into distinct scale space representations, where each wavelet packet (channel) reflected a specific scale and orientation sensitivity Wavelet packet representations for twenty-five natural textures were classified without error by a simple two-layer network classifier An analyzing function of large regularity (D/sub 20/) was shown to be slightly more efficient in representation and discrimination than a similar function with fewer vanishing moments (D/sub 6/) In addition, energy representations computed from the standard wavelet decomposition alone (17 features) provided classification without error for the twenty-five textures included in our study The reliability exhibited by texture signatures based on wavelet packets analysis suggest that the multiresolution properties of such transforms are beneficial for accomplishing segmentation, classification and subtle discrimination of texture >

Interactive aircraft training system and method

Abstract: The system includes a computer (20), a detector (40) for detecting and tracking head orientation and head movement of an individual, a pair of tactile gloves (42, 44) which are donned by the individual (51) for detecting and transmitting to the computer (20) as input data orientation and movements of the hands of the individual inserted in the tactile gloves (42, 44), a stereoscopic, head-mounted display (31), a subsystem for enabling the computer to generate a stereoscopic image of the training environment, a subsystem for causing the stereoscopic image of the training environment to be displayed and changed by the computer relative to input data received by the computer relating to the head orientation and head movement of the individual, relative to input data received by the computer relating to orientation and movement of the hands of the individual inserted in the tactile gloves, and relative to input data reflecting realistic changes in the training environment that would be perceived by the individual if interacting with an identical, non-simulated, training environment. An object (43) representative of a tool is adapted for transmitting tactile information to the computer (20). Sounds representative of the environment are transmitted to the individual (51) through the earphones (33). Vocal emanations of the individual are detected by a microphone (48). System peripheral items (31, 33, 40, 42, 43, 44, 48) are connected to the computer (20) by means of wires (60, 61, 62, 63, 64, 65). The computer (20) interfaces and incorporates an environment simulation and modeling subsystem (22), an image generation component (24), a user interface management component (26), and a component for formulating and transmitting instructional input (28).

Omniview motionless camera orientation system

Abstract: A device for omnidirectional image viewing providing pan-and-tilt orientation, rotation, and magnification within a hemispherical field-of-view that utilizes no moving parts. The imaging device is based on the effect that the image from a fisheye lens (1), which produces a circular image of an entire hemispherical field-of-view, which can be mathematically corrected using high speed electronic circuitry (5, 6, 7, 12 and 13). More specifically, an incoming fisheye image from any image acquisition source (2) is captured in memory (4) of the device, a transformation is performed for the viewing region of interest and viewing direction, and a corrected image is output as a video image signal for viewing, recording, or analysis. As a result, this device can be accomplish the functions of pan, tilt, rotation, and zoom throughout a hemispherical field-of-view without the need for any mechanical mechanisms. The preferred embodiment of the image transformation device can provide corrected images at real-time rates, compatible with standard video equipment. The device can be used for any application where a conventional pan-and-tilt or orientation mechanism might be considered including inspection, monitoring, surveillance, and target acquisition.

Height and gradient from shading

Abstract: The method described here for recovering the shape of a surface from a shaded image can deal with complex, wrinkled surfaces. Integrability can be enforced easily because both surface height and gradient are represented. The robustness of the method stems in part from linearization of the reflectance map about the current estimate of the surface orientation at each picture cell. The new scheme can find an exact solution of a given shape-from-shading problem even though a regularizing term is included. This is a reflection of the fact that shape-from-shading problems are {\it not} ill-posed when boundary conditions are available or when the image contains singular points.

Perceiving spatial relationships in computer-generated images

Abstract: The sources of visual information that must be present to correctly interpret spatial relations in images, the relative importance of different visual information sources with regard to metric judgments of spatial relations in images, and the ways that the task in which the images are used affect the visual information's usefulness are discussed Cue theory, which states that the visual system computes the distances of objects in the environment based on information from the posture of the eyes and from the patterns of light projected onto the retinas by the environment, is presented Three experiments in which the influence of pictorial cues on perceived spatial relations in computer-generated images was assessed are discussed Each experiment examined the accuracy with which subjects matched the position, orientation, and size of a test object with a standard by interactively translating, rotating, and scaling the test object >

An automated method for rotational correction and centering of three-dimensional functional brain images.

Abstract: The display and analysis of functional brain images often benefit from head rotational correction and centering. An automated method was developed to align brain PET images into a standard three-dimensional orientation. The algorithm performs transverse and coronal rotational correction as well as centering of a brain image set. Optimal rotational correction and centering are determined by maximizing a bilateral hemispheric similarity index, the stochastic sign change criterion. Testing of this algorithm on simulated symmetrical brain image sets showed errors less than 1.0 degree and 0.5 pixels for rotational correction and centering, respectively. With actual PET data, the algorithm results correlated well with those obtained by visual inspection. Testing on asymmetrical brain image sets with simulated lesions indicated that performance of the algorithm is not sensitive to focal asymmetries. This automated method provides objective, reproducible image alignment into a standard orientation and facilitates subsequent data analysis techniques for functional brain images.

Visual tracking of known three-dimensional objects

Abstract: A method is described of visually tracking a known three-dimensional object as it moves with six degrees of freedom. The method uses the predicted position of known features on the object to find the features in images from one or more cameras, measures the position of the features in the images, and uses these measurements to update the estimates of position, orientation, linear velocity, and angular velocity of the object model. The features usually used are brightness edges that correspond to markings or the edges of solid objects, although point features can be used. The solution for object position and orientation is a weighted least-squares adjustment that includes filtering over time, which reduces the effects of errors, allows extrapolation over times of missing data, and allows the use of stereo information from multiple-camera images that are not coincident in time. The filtering action is derived so as to be optimum if the acceleration is random. (Alternatively, random torque can be assumed for rotation.) The filter is equivalent to a Kalman filter, but for efficiency it is formulated differently in order to take advantage of the dimensionality of the observations and the state vector which occur in this problem. The method can track accurately with arbitrarily large angular velocities, as long as the angular acceleration (or torque) is small. Results are presented showing the successful tracking of partially obscured objects with clutter.

Translation, rotation, and scale invariant pattern recognition by high-order neural networks and moment classifiers

Abstract: The classification and recognition of two-dimensional patterns independently of their position, orientation, and size by using high-order networks are discussed. A method is introduced for reducing and controlling the number of weights of a third-order network used for invariant pattern recognition. The method leads to economical networks that exhibit high recognition rates for translated, rotated, and scaled, as well as locally distorted, patterns. The performance of these networks at recognizing types and handwritten numerals independently of their position, size, and orientation is compared with and found superior to the performance of a layered feedforward network to which image features extracted by the method of moments are presented as input. >

Stereological measurement and error estimates for three-dimensional fiber orientation

Abstract: A method is presented for measuring three-dimensional fiber orientation in fiber-reinforced polymers and placing confidence limits on the results. The orientations of individual fibers are determined from the elliptical intersections between the cylindrical fibers and a polished section. This can be done using either manual digitization or automated image analysis. Volume averages for the sample are computed using an orientation-dependent weighting function that corrects for the bias of an area-based sample. Equations are developed for nonuniform fiber lengths, using both number-average and weight-average measures of orientation. Sources of systematic, measurement, and sampling error are discussed and equations for sampling error and the propagation of measurement error are derived. The results use a second-rank tensor to characterize fiber orientation, but the error analysis can be applied to any type of orientation parameter. We implement the technique using manual digitization of optical micrographs. Our implementation accurately measures samples with known orientation, and produces identical results from two perpendicular sections of a glass fiber/nylon injection-molded sample.

Statistical-mechanical analysis of self-organization and pattern formation during the development of visual maps

Abstract: We report a detailed analytical and numerical model study of pattern formation during the development of visual maps, namely, the formation of topographic maps and orientation and ocular dominance columns in the striate cortex. Pattern formation is described by a stimulus-driven Markovian process, the self-organizing feature map. This algorithm generates topologically correct maps between a space of (visual) input signals and an array of formal ``neurons,'' which in our model represents the cortex. We define order parameters that are a function of the set of visual stimuli an animal perceives, and we demonstrate that the formation of orientation and ocular dominance columns is the result of a global instability of the retinoptic projection above a critical value of these order parameters. We characterize the spatial structure of the emerging patterns by power spectra, correlation functions, and Gabor transforms, and we compare model predictions with experimental data obtained from the striate cortex of the macaque monkey with optical imaging. Above the critical value of the order parameters the model predicts a lateral segregation of the striate cortex into (i) binocular regions with linear changes in orientation preference, where iso-orientation slabs run perpendicular to the ocular dominance bands, and (ii) monocular regions with low orientation specificity, which contain the singularities of the orientation map. Some of these predictions have already been verified by experiments.

Three-dimensional location estimation of circular features for machine vision

Abstract: A closed-form analytical solution to the problem of 3D estimation of circular-feature location is presented. Two different cases are considered: 3D orientation and position estimation when the radius is known and when it is not known. Extension of the method to general 3D quadratic features is also addressed. Simulated experimental results obtained for all three cases verified the analytical method. In the case of real experiments, a set of circles located on a calibration plate, whose locations were known with respect to a reference frame, were used for camera calibration as well as for the application of the method. A sequential compensation procedure was applied to the input gray-level image to compensate for distortion. These results also showed the validity of the process and the applicability of the analytical method. >

Surface Orientation and Time to Contact from Image Divergence and Deformation

Abstract: This paper describes a novel method to measure the differential invariants of the image velocity field robustly by computing average values from the integral of normal image velocities around image contours. This is equivalent to measuring the temporal changes in the area of a closed contour. This avoids having to recover a dense image velocity field and taking partial derivatives. It also does not require point or line correspondences. Moreover integration provides some immunity to image measurement noise.

Real-time recognition of pointing information from video

Abstract: An occurrence of a predefined object in an image is recognized by receiving image data, convolving the image data with a set of predefined functions to analyze the image data into occurrences of predefined elementary features, and examining the occurrences for an occurrence of a predefined combination of the elementary features that is characteristic of the predefined object. Preferably the image data are convolved directly with a first predefined function to determine blob responses, and a second predefined function to determine ganglia responses indicating edges of objects. Then the ganglia responses are convolved with a third predefined function to determine simple responses indicating lines in the image, and the simple responses are combined with the ganglia responses to determine complex responses indicating terminated line segments in the image. A pointing finger, for example, is recognized from the combination of a blob response and a complex response. The method, for example, permits a data input terminal to recognize in real time the presence, position, and orientation of a pointing finger, to eliminate the need for data input devices such as "mice" or "joysticks." Therefore a user can direct an application program in the most natural way, without the distraction of manipulating a data input device.

Rectification of digital imagery

Abstract: Different methods can be applied to generate digital orthophotos. Three commonly used approaches are analyzed and compared in this paper. They can be applied to rectify both digitized aerial photographs and satellite scenes. These methods are polynomial, projective, and differential rectifications. The first two are defined by analytical transformations between image and orthophoto without considering the geometry and orientation of the camera. They are approximate solutions. The last one models the physical reality of the imaging process by means of the collinearity equations and corrects for relief displacements. All three methods were implemented on a workstation and were tested with digitized aerial photographs and video images. By overlaying GIS data over the digital orthophoto, the quality of the rectification is checked. To determine the planimetric accuracy of the results, the coordinates of targets were measured in a digital orthophoto and compared to known map coordinates.

Measuring method and apparatus

Abstract: The present invention relates to methods and apparatus for measuring the linear dimensions and determining the three-dimensional, spatial volume of objects (620). In the preferred embodiments, ultrasonic sensing is employed, alone or in combination with other measuring methods such as light curtains and the timing (622) of moving objects (620), to ascertain linear dimensions. The preferred embodiments are particularly adapted to the measurement of large and irregularly configured objects, such as palletized loads, and to the measurement of moving, rectangular objects (620) of random orientation with respect to the direction of travel.

Photometric method for determining surface orientation from multiple images

Abstract: A novel technique called photometric stereo is introduced. The idea of photometric stereo is to vary the direction of incident illumi­ nation between successive images, while holding the viewing direction constant. It is shown that this provides sufficient information to deter­ mine surface orientation at each image point. Since the imaging geom­ etry is not changed, the correspondence between image points is known a priori. The technique is photometric because it uses the radiance values recorded at a single image location, in successive views, rather than the relative positions of displaced features. Photometric stereo is used in computer-based image understanding. It can be applied in two ways. First, it is a general technique for deter­ mining surface orientation at each image point. Second, it is a technique for determining object points that have a particular surface orientation. These applications are illustrated using synthesized examples.

Model-Based Object Tracking in Traffic Scenes

Abstract: This contribution addresses the problem of detection and tracking of moving vehicles in image sequences from traffic scenes recorded by a stationary camera. In order to exploit the a priori knowledge about the shape and the physical motion of vehicles in traffic scenes, a parameterized vehicle model is used for an intraframe matching process and a recursive estimator based on a motion model is used for motion estimation. The initial guess about the position and orientation for the models are computed with the help of a clustering approach of moving image features. Shadow edges of the models are taken into account in the matching process. This enables tracking of vehicles under complex illumination conditions and within a small effective field of view. Results on real world traffic scenes are presented and open problems are outlined.

Method and apparatus for detecting and decoding bar code symbols using two-dimensional digital pixel images

Abstract: A system for detecting a bar code symbol receives a two-dimensional image containing the symbol, where the system has no a priori knowledge of the orientation of the symbol in the image. The system scans the image in one dimension to locate a quiet zone for the symbol and searches the image in one dimension to identify a start/stop character of the symbol.

Shape from texture for smooth curved surfaces in perspective projection

Abstract: Projective distortion of surface texture observed in a perspective image can provide direct information about the shape of the underlying surface. Previous theories have generally concerned planar surfaces; this paper presents a systematic analysis of first- and second-order texture distortion cues for the case of a smooth, curved surface. In particular, several kinds of texture gradients are analyzed and are related to surface orientation and surface curvature. The local estimates obtained from these cues can be integrated to obtain a global surface shape, and it is shown that the two surfaces resulting from the well-known tilt ambiguity in the local foreshortening cue typically have qualitatively different shapes. As an example of a practical application of the analysis, a shape-from-texture algorithm based on local orientation-selective filtering is described, and some experimental results are shown.

Color reflectance modeling using a polychromatic laser range sensor

Abstract: A system for simultaneously measuring the 3-D shape and color properties of objects is described. Range data are obtained by triangulation over large volumes of the scene, whereas color components are separated by means of a white laser. Details are given concerning the modeling and the calibration of the system for bidirectional reflectance-distribution functions measurements. A reflection model is used to interpret the data collected with the system in terms of the underlying physical properties of the target. These properties are the diffuse reflectance of the body material, the Fresnel reflectance of the air media interface, and the slope surface roughness of the interface. Experimental results are presented for the extraction of these parameters. By allowing the subtraction of highlights from color images and the compensation for surface orientation, spectral reflectance modeling can help to understand 3-D scenes. A practical example is given where a color and range image is processed to yield uniform regions according to material pigmentation. >

Metric invariance in object recognition: a review and further evidence.

Abstract: Phenomenologically, human shape recognition appears to be invariant with changes of orientation in depth (up to parts occlusion), position in the visual field, and size. Recent versions of template theories (e.g., Ullman, 1989; Lowe, 1987) assume that these invariances are achieved through the application of transformations such as rotation, translation, and scaling of the image so that it can be matched metrically to a stored template. Presumably, such transformations would require time for their execution. We describe recent priming experiments in which the effects of a prior brief presentation of an image on its subsequent recognition are assessed. The results of these experiments indicate that the invariance is complete: The magnitude of visual priming (as distinct from name or basic level concept priming) is not affected by a change in position, size, orienta- tion in depth, or the particular lines and vertices present in the image, as long as representations of the same components can be activated. An implemented seven layer neural network model (Hummel & Bicderman, 1992) that captures these fundamental properties of human object recognition is described. Given a line drawing of an object, the model activates a viewpoint-inva riant structural descrip- tion of the object, specifying its parts and their interrelations. Visual priming is interpreted as a change in the connection weights for the activation of: a) cells, termed geon feature assemblies (GFAs), that conjoin the output of units that represent invariant, independent properties of a single geon and its relations (such as its type, aspect ratio, relations to other geons), or b) a change in the connection weights by which several GFAs activate a cell representing an object.

Stereotactic localization and guidance using a machine vision technique.

Abstract: Machine vision techniques (video cameras) can be used to determine the three-dimensional position of objects. This transformation can be accomplished with standard mathematical algorithms. Initial accuracy tests of stereotactic localization with video cameras were performed using a standard Brown-Roberts-Wells (BRW) phantom simulator coupled with the BRW angiographic localizer. Localization accuracy was within 1.5 mm. Potential applications of machine vision techniques include freehand stereotactic localization of the position and orientation of surgical instruments. With sufficient computer speed these techniques can be used for continuous monitoring of the position of instruments within the cranial vault.

3D relative position and orientation estimation using Kalman filter for robot control

Abstract: A vision-based position sensing system which provides three-dimensional relative position and orientation (pose) of an arbitrary moving object with respect to a camera for a real-time tracking control is studied. Kalman filtering was applied to vision measurements for the implicit solution of the photogrametric equations and to provide significant temporal filtering of the resulting motion parameters resulting in optimal pose estimation. Both computer simulation and real-time experimental results are presented to verify the effectiveness of the Kalman filter approach with large vision measurement noise. >

Ultrasound beam orientation during standard two-dimensional imaging: assessment by three-dimensional echocardiography.

Abstract: Standard two-dimensional echocardiographic image planes are defined by anatomic landmarks and assumptions regarding their orientation when these landmarks are visualized. However, variations of anatomy and technique may invalidate these assumptions and thus limit reproducibility and accuracy of cardiac dimensions recorded from these views. To overcome this problem, we have developed a three-dimensional echocardiograph consisting of a real-time scanner, three-dimensional spatial locater, and personal computer. This system displays the line of intersection of a real-time image and an orthogonal reference image and may be used to assess actual image orientation during standardized two-dimensional imaging when the line-of-intersection display is not observed by the operator. Three hundred forty standard images were assessed from 85 examinations by 11 echocardiographers. Twenty-four percent of the unguided standard images were optimally positioned within ± 5 mm and ± 15 degrees of the standard. Of the optimal images, two thirds were parasternal long-axis views. A subsequent study with three-dimensional echocardiography and line-of-intersection guidance of image positioning showed 80% of the guided images to be optimally positioned, a threefold improvement ( p

Using shadows in finding surface orientations

Abstract: : Given a line drawing from an image with shadow regions identified, the shapes of the shadows can be used to generate constraints on the orientations of the surfaces involved. This paper describes the theory which governs those constraints under orthography. A 'Basic Shadow Problem' is first posed, in which there is a single light source, and a single surface casts a shadow on another (background) surface. There are six parameters to determine: the orientation (2 parameters) for each surface, and the direction of the vector (2 parameters) pointing at the light source. If some set of 3 of these are given in advance, the remaining 3 can then be determined geometrically. The solution method consists of identifying 'illumination surfaces' consisting of illumination vectors, assigning Huffman-Clowes line labels to their edges, and applying the corresponding constraints in gradient space. The analysis is extended to shadows cast by polyhedra and curved surfaces. In both cases, the constraints provided by shadows can be analyzed in a manner analogous to the Basic Shadow Problem. When the shadow falls upon a polyhedron or curved surface, similar techniques apply. The consequences of varying the position and number of light sources are also discussed. Finally, some methods are presented for combining shadow geometry with other gradien space techniques for 3D shape inference. (Author)

Steerable filters and local analysis of image structure

Abstract: : Two paradigms for visual analysis are top-down, starting from high-level models or information about the image, and bottom-up, where little is assumed about the image or objects in it. We explore a local, bottom-up approach to image analysis. We develop operators to identify and classify image junctions, which contain important visual cues for identifying occlusion, transparency, and surface bends. Like the human visual system, we begin with the application of linear filters which are oriented in all possible directions. We develop an efficient way to create an oriented filter of arbitrary orientation by describing it as a linear combination of basis filters. This approach to oriented filtering, which we call steerable filters, offers advantages for analysis as well as computation. We design a variety of steerable filters, including steerable quadrature pairs, which measure local energy. We show applications of these filters in orientation and texture analysis, and image representation and enhancement. We develop methods based on steerable filters to study structures such as contours and junctions. We describe how to post-filter the energy measures in order to more efficiently analyze structures with multiple orientations. We introduce a new detector for contours, based on energy local maxima. We analyze contour phases at energy local maxima, and compare the results with the prediction of a simple model. Using these tools, we analyze junctions. Based on local oriented filters, we develop simple mechanisms which respond selectively to T, L, and X junctions. The T and X junctions may indicate occlusion and transparency, respectively. These mechanism show that detectors for important, low-level visual cues can be built out of oriented filters and energy measures, which resemble responses found in the visual cortex. We present a second approach to junction detection based on salient contours.

Helicopter virtual image display system incorporating structural outlines

Abstract: A virtual image display system provides video displays based upon virtual images of the external world having synchronized structural outlines superimposed on the video displays to a pilot operating an aircraft such as a helicopter in non-visual flight conditions. The virtual image display system includes a virtual imaging subsystem for generating virtual images of the external world, a video display subsystem for generating video images based upon the virtual images and for displaying the video images for the pilot's viewing, a sensing means for providing signals corresponding to the spatial location and perspective of the video display subsystem, a map comprising structural outlines corresponding to structural members forming the canopy structure of the helicopter, and a computer subsystem providing electronic interfacing between the elements of the virtual image display system, for synchronizing the orientation of the virtual imaging subsystem with the video display subsystem, and for processing the virtual image signals to provide signals to the video display subsystem to generate video displays. The computer utilizes the helmet position signals to define the relative position and orientation of the video display subsystem in the cockpit, and utilizes such definition to reconstruct structural outlines from the map that are synchronized to the perspective of the video images. The synchronized structural images are superimposed upon the video display.