Search

In this paper, we develop an improved algorithm to recover and correct polyhedron from a given line drawing. It can immediately recover the polyhedron even if image data are incorrect due to vertex-position errors. Many experiments have been done to demonstrate the algorithm to be effective and efficient and show that the speed of recovery and correction is greatly improved and one of experiments is given.

In this paper, a reconstruction method of a curved object from a single gray-level image is proposed. To realize easy but robust shape from shading (SFS), the proposed method adopts isodensity lines. The results of computer simulation using virtual objects show that the prospects of using this method are very encouraging.

D shape measurement has emerged as a very useful tool in numerous fields because of its wide and ever-increasing applications. In this paper, we present a passive, fast and accurate 3D shape measurement technique using stereo vision approach. The technique first employs a scale-invariant feature transform algorithm to detect point matches at a number of discrete locations despite the discontinuities in the images. Then an automated image registration algorithm is applied to find full-field point matches with subpixel accuracy. After that, the 3D shapes of the objects can be reconstructed according to the obtained point matching and the camera information. The proposed technique is capable of performing a full-field 3D shape measurement with high accuracy even in the presence of discontinuities and multiple separate regions. The validity is verified by experiments.

The ability of younger and older adults to perceive the 3-D shape, depth, and curvature of smooth surfaces defined by differential motion and binocular disparity was evaluated in six experiments. The number of points defining the surfaces and their spatial and temporal correspondences were manipulated. For stereoscopic sinusoidal surfaces, the spatial frequency of the corrugations was also varied. For surfaces defined by motion, the lifetimes of the individual points in the patterns were varied, and comparisons were made between the perception of surfaces defined by points and that of more ecologically valid textured surfaces. In all experiments, the older observers were less sensitive to the depths and curvatures of the surfaces, although the deficits were much larger for motion-defined surfaces. The results demonstrate that older adults can extract depth and shape from optical patterns containing only differential motion or binocular disparity, but these abilities are often manifested at reduced levels of performance.

Last year we presented the influence of pose on threedimensional (3-D) shape classification in the context of a repeatability study. Meaning, that the subjects are repeatedly scanned 10 times and they attempt to assume the same pose each time. It was shown that changes due to a slight pose modification had no detrimental effects for shape classification. This paper discuss a second set of experiments, designed to test the stability of the geometric search engine in more extreme cases. The pose of the subjects between the scans are modified substantially, with an increasing amount of differences compared to the CAESAR protocol. Experimental results will be presented and discussed., Proceedings of the Digital Human Modeling for Design and Engineering Conference, June 26-28, 2001., Arlington, Virginia, USA

Proceedings of the Digital Human Modeling for Design Engineering, June 6-8, 2000., Dearborn, Michigan, USA

視覚センサを用いて物体の3次元形状を識別する場合, システムは物体の移動および回転の影響を受けにくいことが要求される.筆者らは, 輝点マトリクス投影による空間のコード化および画像データのパターン変換を用いた効率の良い3次元形状識別法を開発した.

In this paper we describe a system that is able to acquire models of 2-D shapes from cluttered scenes. The input of the system is a sequence of images each of which shows an unknown number of overlapping unknown 2-D objects. The system identifies matching partial shapes across different images and combines them into complete 2-D shape models thus giving a complete interpretation of the input scenes. The identification of partial shapes is based on string matching, whereas a graph search procedure is used for shape model generation. The system has been fully implemented and tested on images containing parts of a jigsaw puzzle.

Two observations are worthy of note. First, the experienced optical engineer can usually determine, sometimes partly subjectively, the positions of phase wraps in the image. This suggests that the information necessary to identify phase wraps does exist. Second, to date, no universally applicable technique for phase wrap detection is available. Indeed, it may be that there is no straight forward analytical method that can be used. The concept of a neuron was first postulated by McCulloch and Pitts (1943) and a neural network provides a mechanism by which a machine can learn from experience. The foregoing discussion suggests that neural network technology may be suitable for addressing the phase unwrapping problem. This paper describes preliminary work on the use of neural networks to identify phase wraps earlier in the phase measuring process, prior to the calculation of wrapped phase.

This paper gives an overview of 3-D shape measurement techniques, especially those developed in our laboratory, and their applications. These techniques include moire topography(MT), Fourier transform profilometry (FTP), phase measuring profilometry (PMP), and 360° shape measurement.

In images of textured 3-D surfaces, orientation flows created by the texture components parallel to the surface slant play a critical role in conveying the surface slant and shape. This study examines the visibility of these orientation flows in complex patterns. Specifically, we examine the effect of orientation of neighboring texture components on orientation flow visibility. Complex plaids consisting of gratings equally spaced in orientation were mapped onto planar and curved surfaces. The visibility of the component that creates the orientation flows was quantified by measuring its contrast threshold (CT) while varying the combination of neighboring components present in the pattern. CTs were consistently lowest only when components closest in orientation to that of the orientation flows were subtracted from the pattern. This finding suggests that a previously reported frequency-selective cross-orientation suppression mechanism involved with the perception of 3-D shape from texture is affected by proximity in orientation of concurrent texture components.

Surface contamination of insulators in high voltage transmission and distribution systems may lead to troublesome flashovers which interrupt service. Insulator contamination may be monitored using direct measurement of equivalent salt deposit density (ESDD), which directly relates to the flashover voltage. A fiber-optic evanescent- coupling fiber-optic ESDD monitor is presented. The use of a piece of D-shape E-core fiber as sensor head can provide a sensitive and large area ESDD monitoring system. We have demonstrated the use of D-shape fiber in monitoring ESDD on insulator surfaces. A simple process has been developed to precisely etch the D-fiber. The polarimetric and evanescent loss sensors have been investigated. The evanescent loss sensor is particularly suitable for in-situ ESDD monitoring in power transmission lines and substations.© (1996) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

In the development of ESK(Experimental SMES of Kyushu electric power, 1kWh/1MW) which is a toroid type coil composed of 6 elementary coils wound by strand cables, problems of mechanical force mat might be connected with mechanical disturbance in coils are very important. To reduce this force, a new coil shape is proposed named “modified D shape” that has two straight parts mechanically connected to each other by support material as well as two curved parts having the shape with no bending moment. It is made clear that the reduction of floor area(2/3) as well as that of tension (1/2) without the increase of amount of conductor are shown to be attained by calculation.

Standard input tools such as the mouse and keyboard do not provide users with a direct and intuitive means of 3-D shape manipulation. This study proposes a new concept of interface system for 3-D shape deformation using a deformable real object as an input tool. By deforming the tool with bare hands with a tactile feedback, users can manipulate a 3-D shape modeled and displayed on a computer screen directly and intuitively. A prototype with a cubical input tool made of electrically conductive polyurethane foam demonstrates the effectiveness and promise of the concept.

Standard input devices such as a mouse and a keyboard in present computer-aided design systems do not provide users with direct and intuitive facilities for highly 3-D shape manipulation. To solve the problem, this paper proposes a new interface system for 3-D shape manipulation by adopting a real elastic object as an input device. By deforming the device with bare hands with a tactile feedback, users can manipulate a 3-D shape modeled and displayed on a computer screen quite directly and intuitively. A prototype with a cubical input device made of electrically conductive polyurethane foam is also presented.

Neural architectures may offer an adequate way to deal with early vision since they are able to learn shape features or classify unknown shapes, generalising the features of a few meaningful examples, with a low computational cost after the training phase. Two different neural approaches are proposed by the authors: the first one consists of a cascaded architecture made up by a first stage named BWE (Boundary Webs Extractor) which is aimed to extract a brightness gradient map from the image, followed by a backpropagation network that estimates the geometric parameters of the object parts present in the perceived scene. The second approach is based on the extraction of the boundary webs map from the image and its comparison with boundary webs maps exhaustively generated from synthetic superquadrics. A purposely defined error figure has been used to find the best match between the two kinds of maps. A functional comparison between the two systems is described and the quite satisfactory experimental results are presented.

A novel dual-frequency pattern is developed which combines a high-frequency sinusoid component with a unit-frequency sinusoid component, where the high-frequency component is used to generate robust phase information, and the unit-frequency component is used to reduce phase unwrapping ambiguities. With our proposed pattern scheme, phase unwrapping can overcome the major shortcomings of conventional spatial phase unwrapping: phase jumping and discontinuities. Compared with conventional temporal phase unwrapped approaches, the proposed pattern scheme can achieve higher quality phase data using a less number of patterns. To process data in real time, we also propose and develop look-up table based fast and accurate algorithms for phase generation and 3-D reconstruction. Those fast algorithms can be applied to our pattern scheme as well as traditional phase measuring profilometry. For a 640 x 480 video stream, we can generate phase data at 1063.8 frames per second and full 3-D coordinate point clouds at 8.3 frames per second. These achievements are 25 and 10 times faster than previously reported studies.

The Explicit Approximation Method for 2D shape optimization under displacement constraint is presented. This is a zeroth order sequential approximation method. The basic idea is to use mechanical knowledge instead of purely numerical gradient information to form an explicit approximation to the implicit constraint. Design examples are presented to show that the displacement constraint usually has a good explicit approximation in 2D problems, and a good and feasible, if not theoretically optimum, design can be found very efficiently by the Explicit Approximation Method using only zeroth order information.

This paper presents a new afgorithm for smoothly blending between two 2-D polygonal shapes. The algorithm is based on a physical model wherein one of the shapes is considered to be constructed of wire, and a solution is found whereby the first shape can be bent and/or stretched into the second shape with a minimum amount of work. The resulting solution tends to associate regions on the two shapes which look alike. If the two polYgons have m and n vertices respectively, the afgorithm is O(mn). The algorithm avoids local shape inversions in whkh intermediate polygons self-intersect, if such a solution exists.

A new method of 3-D shape measurement by moire technique is presented. An interferometer and computer were used instead of gratings to obtain moire contour. The principle of it is theoretically analyzed in detail. In order to obtain moire contour without unwanted patterns, we propose a method of obtaining moire contour by XOR logic using a computer. Experimental results are given.

A method based on differential geometry is presented for quantitatively measuring facial (or body) surfaces and the amount of change produced in them by facial surgery. Facial shape is described mathematically in terms of 8 fundamental surface types. Changes produced in the shape of the face are described in terms of changes in these constituent surfaces. The method is illustrated using examples of faces changed by reconstructive surgery and its relationship to clinical observation explained.© (1991) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

In this paper a new approach is proposed for industrial part recognition. The basic ideal of the method is that the 3D geometry parameters of parts are combined with their shape invariant features for recognition. An effective method is presented in this paper to extract the approximate 3D information of the parts from an image and a machine vision system for recognizing flat industrial parts based on the approach has been developped on PCVISION image processing system. The experimental results show that the approach works well for recognizing shape distorted flat parts.© (1990) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Various biomedical bbjects are 3D curveshaped. The paper discusses various methods a-F recovering the shape a-F such objects from multiframes under various geometrical and physical settings. 1.© (1990) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

This paper presents parallel computable contour based feature strings for two dimensional shape recognition. Parallel techniques are used for contour extraction and for computation of normalized contour based features independent of scale and rotation. The EREW PRAM architecture is considered, but the technique can be adapted to other parallel architectures. Illustrated examples are presented.© (1990) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

In this paper, the framework of a 3-D shape reconstruction software system used in biomedical research is reported. The system mainly includes two packeges: 2-D contours extraction and 3-D shape reconstruction. The new methods used in this system are described. This paper also shows the application of the system in neuro-physiology research and medical CT images.

The latest time-frequency tool, wavelet analysis, is presented to extract the phase information φ ( x , y ) from a single spatial carrier fringe pattern. First, the fringe pattern is generated and projected onto an object. The spatial carrier fringe pattern, carrying the 3-D shape of the object, is then captured using a CCD camera. Afterward, the spatial carrier fringe pattern is modulated by two high-frequency waves, which makes the spectrum of the carried phase relatively lower among the spectra of two modulated patterns. Following this, the spectrum of the object is separated into the approximation subband using a wavelet transform. The phase map of the 3-D shape of the object is then extracted from the mixed spectra by reconstructing only the approximation of their wavelet transforms. Finally, the 3-D shape of the object can be calculated from the phase map.

An optical method for shape and deformation measurement of a rotating blade, which is based on Fourier transform profilometry FTP and the stroboscopic effect, is proposed and has been verified by experiments. The integrated principles of the method, the process of 3-D phase unwrapping, and the detailed design of the synchronization con- trol unit are discussed. Experimental results on a commercial electric fan prove that FTP can reconstruct the shape of a rotating blade efficiently and reveal its deformation different times clearly by using stroboscopic structured illumination. This method can be used for studying high-speed motion with rapid periodic signals. © 2005 Society of Photo-Optical Instrumentation Engineers. DOI: 10.1117/1.2127927 Subject terms: Fourier transform profilometry; 3-D shape measurement; stroboscope; dynamic 3-D visualization; structured illumination; phase unwrapping; rotating blade.

A new concept for 3-D shape measurement of complex ob- jects is proposed, which is based on photogrammetry and fringe projec- tion. The shape of a partial view is determined by a combination of the phase-shift method for fringe evaluation and a photogrammetric triangu- lation to calculate the 3-D coordinates related to the sensor coordinate system. For the measurement of complex objects, partial views are taken from different sensor positions. The problem of matching the par- tial views into each other is solved by transforming each sensor position and the relating point cloud of the shape in a global coordinate system by photogrammetric matching of reference targets. © 2000 Society of Photo- Optical Instrumentation Engineers. (S0091-3286(00)02701-X)

Flat, rectangular displays consisting of 2, 3, 4, or 6 vertical grey stripes, alternately light and dark gray (NCS 3500 and NCS 6000, respectively) were presented one at a time illuminated by two identical projectors. One projector illuminated the upper half of the display through a yellowish colour filter (Strand Golden Amber), and the other one illuminated the lower half through a bluish colour filter (Kodak Wratten 80 A). The display appeared ambiguous, periodically and spontaneously shifting between two distinct 3-D shape percepts, A and B. Display A appeared vertically folded along the reflectance edges (‘shape from shading’, where the dark fields appeared to be attached shadows). The colours were quite saturated yellow and blue surface colours. There was no colour constancy; Display B appeared horizontally folded along the illumination edge like a roof (‘shape from shading’, where one of the illuminations appeared to be an attached shadow). The display now appeared very desaturated, even achromatic, but in a ‘warm’ illumination. The colour constancy was almost complete. The phenomenon is demonstrated with a slide, and some psychophysical data on the colour desaturation and on the frequency of shifts between the two percepts are reported as well as some observations on combinations of illuminant colours other than yellow and blue. The reported colour constancy phenomenon is discussed in relation to an earlier presented model for the perception of illumination, colour, and depth [S S Bergström, 1994, in Lightness, Brightness, and Transparency Ed. A Gilchrist (Hillsdale, NJ: Lawrence Erlbaum Associates)].