In this paper, a new measuring system is proposed which can measure the 3D position and orientation of rigid bodies. Its measurement principle is based on detection of laser beam reflected from a specially fabricated mirror that looks like a triangular pyramid having an equilateral cross-sectional shape. The mirror has three lateral reflective surfaces inclined 45 degree(s) to its bottom surface. We call this mirror 3-facet mirror. The 3-facet mirror is mounted on the object whose position and orientation are to be measured. The measurement is operated by a laser-based optical system composed of a 3-facet mirror, a He-Ne laser source, three position-sensitive detectors.

A system of small rotation angle measurement based on the fringe projection is proposed and demonstrated. This system has potential for a broad range of uses and a robustness for the external disturbances, because it requires no coherent light. The setup is very simple and applicable to the automatic on-line measurement. Several measurements indicate a sensitivity of 3 arcsec.

A new method for measuring complex refractive index is presented based on Fresnel's equations and the uses of a lock-in amplifier. A lock-in amplifier is introduced into a common path interferometer to measure the product of the amplitude reflection coefficients of s and p polarizations and their corresponding phase difference of a light beam reflected from a medium with complex refractive index. Then, these data are substituted into the special equations derived from Fresnel's equations, the complex refractive index can be obtained by numerical calculations. The resolution of this method is better than 0.01. It has both merits of a conventional common-path interferometer and a heterodyne interferometer. And its validity is demonstrated.

A novel method for measuring small displacements based on the total-internal-reflection-heterodyne interferometry is presented. In this method, a common-path heterodyne interferometer is applied to measure the phase differences between s and p polarizations at total-internal reflection. The phase differences depend on the incident angles which are the function of displacements by using the image formula. Hence, small displacements can be evaluated only by measuring the phase differences. In our optical setup, a heterodyne light source with a frequency difference 800 Hz between s and p polarizations is used. A light beam passing through a lens is reflected from a mirror located near the focal plane of the objective lens and is driven by a piezo- electric transducer. If the mirror shifts from the focal plane of the lens, the reflected beam deflects with a small angular deviation to a phase difference measurement system. From Fresnel's equations, the phase difference between s and p polarizations is the function of incident angle and the numbers of TIR at total-internal reflection condition. To compare this test signal with a reference signal using a phase meter, the phase difference can be measured in real- time. Consequently, the displacement also can be calculated.

The criteria were determined for simultaneous observation of scattered and unscattered (ballistic) peaks in temporal distribution of a narrow pulsed laser beam passed through a strongly scattering medium. Conditions were found for such an observation on the basis of non-stationary two-flux model for radiation transport. Calculation results permitted to define a compromise between parameters of the initial laser pulse and properties of a scattering medium. This has provided the separate observation of various types of photons. The theoretical calculations were confirmed by experiments on recording pulse profiles of passed radiation of fs-laser in a water-milk solution.

This paper presents a state of the art of reverse engineering for rapid product development. This domain concerns both 3D measurement for comparison between the CAD reference model and the discrete data (point clouds, CT contours), and data acquisition for CAD modeling (surface and volume modeling from grids of set of points or contours). This paper presents the main aspects of reverse engineering domain, from the sensors used for data acquisition to software for data computation and to software for part modeling from point clouds.

This paper introduces some results of a research work carried out on the automation of digitizing process of complex part using a precision 3D-laser sensor. Indeed, most of the operations are generally still manual to perform digitalization. In fact, redundancies, lacks or forgetting in point acquisition are possible. Moreover, digitization time of a part, i.e. immobilization of the machine, is thus not optimized overall. So, it is important, for time- compression during product development, to minimize time consuming of reverse engineering step. A new way to scan automatically a complex 3D part is presented to order to measure and to compare the acquired data with the reference CAD model. After introducing digitization, the environment used for the experiments is presented, based on a CMM machine and a plane laser sensor. Then the proposed strategy is introduced for the adaptation of this environment to a robotic CAD software in order to be able to simulate and validate 3D-laser-scanning paths. The CAPP (Computer Aided Process Planning) system used for the automatic generation of the laser scanning process is also presented.

We present a framework for estimating 3D relative structure (shape) and motion given objects undergoing non-rigid deformation as observed from a fixed camera, under perspective projection. Deforming surfaces are approximated as piece-wise planar, and piece-wise rigid. Robust registration methods allow tracking of corresponding image patches from view to view and recovery of 3D shape despite occlusions, discontinuities, and varying illumination conditions. Many relatively small planar/rigid image patch trackers are scattered throughout the image; resulting estimates of structure and motion at each patch are combined over local neighborhoods via an oriented particle systems formulation. Preliminary experiments have been conducted on real image sequences of deforming objects and on synthetic sequences where ground truth is known.

A unique three camera stereoscopic omnidirectional viewing system based on the periscopic panoramic camera described in the 11/98 SPIE proceedings (AM13). The 3 panoramic cameras are equilaterally combined so each leg of the triangle approximates the human inter-ocular spacing allowing each panoramic camera to view 240 degree(s) of the panoramic scene, the most counter clockwise 120 degree(s) being the left eye field and the other 120 degree(s) segment being the right eye field. Field definition may be by green/red filtration or time discrimination of the video signal. In the first instance a 2 color spectacle is used in viewing the display or in the 2nd instance LCD goggles are used to differentiate the R/L fields. Radially scanned vidicons or re-mapped CCDs may be used. The display consists of three vertically stacked 120 degree(s) segments of the panoramic field of view with 2 fields/frame. Field A being the left eye display and Field B the right eye display.

Uncompensated thermomechanical errors in laser tracking interferometers are examined by evaluating the difference between tracking interferometer compensations in a controlled laboratory environment versus being compensated in a factory environment. The hypothesis under test was that compensation in a factory environment does not adversely affect, and may actually improve, the uncertainty of laser tracker systems. This hypothesis was confirmed by measuring a standard (i.e., linear interferometer) using laboratory- compensated and certified instruments, and then compensating the instrument in the factory environment and re-measuring the standard. The results showed that in-shop compensation generates less variation in the measurement of the standard when compared to the laboratory-compensated and certified instruments. Certified weather stations are used to compensate for the uncontrolled atmospheric effects on the range measurement.

This paper describes the analysis and identification of road surfaces 3D distress characteristics by means of a three- line laser sensor. A laser-based vehicle-mounted imaging system simultaneously evaluates several conditions of road surface distress, such as ruts, cracks, and surface roughness. The 3D feature extraction is done by a combination of two measuring techniques: triangulation and defocusing. A generated pattern of transverse profiles spaced by 11 cm, is used to sample the road surface. Physical differences between profiles (beam shaping, back- scattering angles, etc.) require special signal processing and analysis techniques in order to reconstruct a 3D image. A new fuzzy system is applied to the analysis and identification of surface cracks. The fuzzy reasoning takes into account the vagueness and the subjectivity of the classification process provided by human experts in the formulation of the identification criteria. This approach is based on a limited number of linguistic fuzzy rules given in natural language, which are extracted from the experimental description of road surface characteristics. Based on a limited number of dispersed events, the fuzzy system is able to predict the evolution of the cracking on the road surface. Theoretical aspects and experimental results are presented.

When range cameras are used for analyzing irregular material on a conveyor belt there will be complications like missing segments caused by occlusion. Also, a number of range discontinuities will be present. In a frame work towards stochastic geometry, conditions are found for the cases when range discontinuities take place. The test objects in this paper are pellets for the steel industry. An illuminating laser plane will give range discontinuities at the edges of each individual object. These discontinuities are used to detect and measure the chord created by the intersection of the laser plane and the object. From the measured chords we derive the average diameter and its variance. An improved method is to use a pair of parallel illuminating light planes to extract two chords. The estimation error for this method is not larger than the natural shape fluctuations (the difference in diameter) for the pellets. The laser- camera optronics is sensitive enough both for material on a conveyor belt and free falling material leaving the conveyor.

An electronic method is proposed that can simultaneously measure an objects shape and its displacements in the micrometer range as well as in the millimeter range. This opens up the possibility to easily integrate holographic deformation measurements with FEA-modeling. It also makes it possible to record a 3D-movie. The method is a combination of digital holography and light-in-flight recording by holography. A Twyman-Green type interferometer is used with the reference mirror replaced by a Littrow-mounted reflection grating. The grating creates an optical delay across the beam profile. The deformations are projected to the 3D-shape whose gradients are determined by Sobel- operators. Large rigid body displacements are calculated by crosscorrelation of shapes measured before and after displacement. A short theoretical description is followed by experimental results of this method.

A light beam structured as a hollow cone is obtained by oblique illumination of a cylindrical surface with a laser beam. The cylindrical surface may be a metallic needle or the core of an optical fiber. In the first case a speckle pattern appears around the light cone. The direction and mean length of light speckles are studied in the text. In the second case the light intensity distribution around the light cone presents bright and dark regions resulting from interference between light rays reflected from and passing through the fiber core. The hollow light cone is transformed into a hollow light cylinder (HLC) by means of a lens whose focus coincides with the cone vertex. The HLC is used to explore the cylindrical layer placed in the neighborhood of the internal wall of a combustion chamber model. Due to the large diameter of the model a Fresnel lens is used to transform the light cone into a light cylinder. This introduces additional optical noise in the system. The HLC is collected again by another Fresnel lens at the model output. A photodetector placed at the focus of the collected beam produces an electrical output showing pulses whenever residual particles pass through the explored region. A quantitative experimental analysis of the performance of the combustion chamber is thus performed.

Several examples of dynamic objects 3D images and high-speed process observations and measurement, are presented. The gun-powder burning process. The real droplet distribution in diesel fuel sprayers. The pressure distribution in T-joint of automobile fuel discharge system. The thin wires explosions--the model of real bursting processes. The pulsed discharge lamps, using for solid-state lasers pumping. The gas-liquid flows in tubes. Measuring the velocity distribution in waterfall at hydra-electric station model. All these and alike measurements at Holography Lab of VNIIOFI, (Moscow, Russia) for different scientific and industrial establishments, were provided.

A new 3D surface contouring and ranging system based on a digital fringe projection and phase shifting technique is described. In this system, three phase-shifted fringe patterns and a centerline pattern are used to determine the absolute phase map of the object. This phase map is then converted to the absolute x, y, and z coordinates of the object surface by a transformation algorithm. To determine the accurate values of the system parameters as required by the transformation algorithm, a two-step calibration procedure was developed. First the parameters were indirectly measured through experiments to determine their approximate values. Second, a calibration plate whose features were calibrated by a coordinate measuring machine was measured by the system at various positions. An iteration algorithm was then used to estimate the system parameters. Measurements of the calibration plate, a sheet- metal panel, and a Ford master gauge showed results consistent with the actual surface contours of the objects.

3D measurements of industrial surfaces is increasingly needed to meet today's quality control requirements, especially in the semiconductor industry. Camera based moire interferometry systems employing phase shifting analysis offer high resolution, but have not commonly been used because of limited data acquisition rates. This paper presents a technique called Scanning Moire Interferometry (SMI^TM) which provides the advantage of high resolution offered by phase shifting analysis while acquiring data at megaHertz rates. The technique is based on the use of a tri- linear CCD sensor. The moire light pattern projector is fixed in position with respect to the tri-linear sensor versus traditional techniques which shift the pattern projector with respect to the sensor between image acquisitions. The SMI^TM technique continuously moves the target with respect to the sensor/projector pair. In this configuration the tri-linear sensor acquires three images, each of which represent a difference phase shift of the projected line pattern. Traditional phase measurement techniques are used to derive a height map of the target from the three phase shifted images. A sensitivity analysis and height measurements of a ball grid array semiconductor package are presented. The advantages of the SMI^TM technique are high resolution, high speed and no moving parts in the sensor assembly.

A variable resolution two color video moire system has been built which uses a dual laser illuminated Mach-Zehnder interferometer as the fringe projector. One input of the interferometer is illuminated by a He-Ne laser and the other by an Argon-Ion laser, allowing simultaneous projection of red and green fringes with difference spatial frequencies. The dual interferometer outputs are also used, allowing illumination of both the target and reference surfaces. Video cameras view the two surfaces and by mixing the video signals with a chroma-key mix in a video special effects generator, real time moire can be observed. A piezo- electrically driven mirror can be used to continuously shift the fringes while taking a video average, which eliminates the fringes in the final image and improves the signal-to- noise of the moire contours. If the reference surface is a plane, the moire contours seen are equal depth contours with higher than normal resolution due to the two fringe systems. Because the moire contours or the distorted fringes in each color can be independently recovered, improved shape recovery other discontinuities is possible. By a different configuration of the projection optics following the interferometer, it is possible to spatially separate the red and green projected fringes and simultaneously illuminate two different shaped reference surfaces. When operated in this mode with flat plate and perfect object reference surfaces, both conventional equal depth moire and error map moire can be seen, one in each color. Applications and limitations of the system will be discussed.

Close range photogrammetry for industrial applications is a demanding field of research. There are several groups in the world working towards a measuring system for quality control applications. Industrialists have realized that a picture is worth than one thousand measures and nowadays more and more off the shelf systems can be found in industrial facilities. However a fully automated vision-base 3D measurement robot is a complex system containing many components which constrain one another. Problems like solving the epipolar geometry in a non-structured environment, especially in real time applications, is still a very hard task. In this article a new laser system which solves the problems of correspondence and occlusions is presented. This laser scanner generates dynamically a grid of a large amount of lines. The projection lines are measured with subpixel precision with a binocular stereo mount. A very fast algorithm for detecting the intersection of lines with subpixel accuracy is described. Furthermore an interesting way for solving the correspondences between the two images is explained. Based on the fact that correspondence between the camera is solved, autocalibration can be performed on line.

Structured light systems made to provide dense data over full image fields present a unique challenge to the task of calibration. Localized artifacts made for CMM or hard gages are often a poor fit for testing actual 3D performance. This paper considers the use of a sine wave artifact to provide a mapping of a calibration matched to full-field capabilities. The sine wave offers the advantages of a continuous function across the full field, with a well defined and easy to analyze shape. Changes in scale in all dimensions, as well as localized variations can be mapped in clear detail using this method.

With the increasing improvement on painting techniques in automotive industry, high quality paint appearance with high gloss, clear image and low orange peel can be achieved on car bodies. In this case, another defect called `metal texture effect' is becoming more significant to paint appearance. In this paper, image texture analysis method, gray-tone spatial-dependence matrices have been employed to evaluate metal texture effect on paint appearance. The preliminary experimental result showed that this method could effectively evaluated metal texture effect as perceived with human eye.

At first, this paper introduces the principle of our section outline sensor using the single-stripe pattern. At second, the measurement error of our sensor is analyzed: the error factors causing the measurement error of our sensor are found out; the mathematical expressions of the measurements are deduced. At third, the measures to reduce the measurement error and the design rules of our sensor are given. At last, a design example is given.

We have continued research on our 3D shape measurement system, using spatial projections of light variation patterns. This method is non-contact, non-invasive, and completes measurement in a short time. However, light diffusion influence on the measurable accuracy. We proposed a method using differentiation to enhance accuracy of measurement, and were successful in achieving this goal. Moreover, the measurable area was expanded. Since our method involves computational manipulation of the data obtained by the original system, it requires no additional equipment. It is a very practical and effective method.