This paper addresses the implementation of sensory basis to support enhanced teleoperation. By adopting multiple structured light systems, each of which consists of a camera and patterned beam projector, the sensor basis supposedly provides an added level of flexibility and adaptability to variable and unstructured task conditions. The technological innovations include calibration of multiple structured light systems and on-line 3D pose measurement. The application of sensory basis is demonstrated for two types of enhanced teleoperation, namely tele-autonomy and tele-collaboration. The sensory information aids in tele-autonomy by generating autonomous motions, and is utilized in tele-collaboration for generating virtual fixtures.

Vision-based motion tracking is commonly used in surveillance, human-machine interfaces in robotics and automation, virtual and augmented reality applications and biomechanics. Most techniques require markers, use a predefined motion sequence or user-intervention for initialization, and do not process in real-time. This paper describes the implementation of a vision-based non-invasive technique for markerless real-time tracking of human-arm motion. Human-arm motion is tracked by processing images from two calibrated cameras in real-time to estimate the position of the 3D joint centers of the wrist and elbow, and determine the orientation of the hand from the 3D positions of the index finger and thumb. Tracking of the hand and arm was carried out without any prior knowledge of subject's arm length, texture, width and distance from the camera.

Mathematical basis for the digitization of gray value of the image is proposed. This can be called Oteru-Kochimizu's Quantization Theorem (OK-QT), just in the same meaning of Shannon's Sampling Theorem (Shannon-ST) for the digitization of the shape of the image. Inspired by the fact ST is the reconstruction theorem of the analog image from discrete image, OK-ST was modeled as the reconstruction theorem of the shape of the probability density function of gray value of the image. This is the novel and unique mathematical basis for the digitization of the gray scale of the image. This paper shows the outline of this theorem and shows also some experimental results to demonstrate its practical applicability. Through this, OK-QT gives a clue to the mathematical paradigm for the complete digitization basis, together with Shannon ST.

An efficient algorithm for searching similar images from databases of a large set of images is proposed. For designing the algorithm, all the correlation coefficient values of registered images are calculated in advance to online computation and they are memorized as a set of keys for efficient search. We theoretically derive an interval estimation of any correlation coefficient between an object image and arbitrary registered images in terms of a pivot image that is one of candidates of the unique solution image and can be simply selected. Using the interval estimations on all the other registered images, some conditions for selecting redundant images from the registered images can be derived and evaluated and then those who has any smaller similarity than the one computed by the pivot image can be skipped away from correlation computation, which efficiently enables to save a lot of computational cost for search. The algorithm was applied to real searching problems in an image database of 1,200 images taken from the real world and to search in multiple template matching problems, for example rotation invariant matching and a search problem on binary maps, resulting the efficiency of the proposed method for the real problems.

Real-time face and iris detection on video sequences has been used to study the eye function and in diverse applications such as drowsiness detection, virtual keyboard interfaces, face recognition and multimedia retrieval. A non-invasive real time iris detection method was developed and consists of three stages: coarse face detection, fine face detection and iris detection. Anthropometric templates are used in these three stages. Elliptical templates are used to locate the coarse face center. A set of anthropometric templates which are probabilistic maps for the eyebrows, nose and mouth are used to perform the fine face detection. Face rotations are considered by rotating the anthropometric templates in fixed angles in steps of 10 degrees. Iris position is then determined within the eye region using another template with concentric semi-circles to compute a line integral in the boundary iris-sclera. The position with the maximum value indicates the iris center. The new method was applied on 10 video sequences, with a total of 6470 frames, from different people rotating their faces in the coronal axis. Results of correct face detection on 8 video sequences was 100%, one reached 99.9% and one 98.2%. Results on correct iris detection are above 96% in 9 of the video sequences and one reached 77.8%. The method was implemented in real-time (30 frames per second) with a PC 1.8 GHz.

In order to gain generality, robustness and efficiency in search, a novel search is proposed based on a representation called `Depth aspect image' is proposed as a controllable two-dimensional representation of local depth distribution used in cooperation with a distinct `Voxel framing', which enables effective reference coordination without any prominent features, such as vertices or edges. A robust statistical estimator called `Least quantile of residuals' is furthermore introduced for robust matching, which can be utilized for both depth matching and model verification. Since the proposed method is of model-based approach with possible views of local structures, the computation cost for matching has to be reduced by introducing random sampling and an effective hashing. Experiments with real scenes show the effectiveness of the proposed method.

The fringe pattern correlator (FPC) can recognize 3D objects using a height transformed complex amplitude as 3D information. The height transformed complex amplitude is a two-dimensional (2D) complex amplitude whose phase factor includes actual height information of 3D objects. The FPC can perform effective 3D correlation using 2D correlation with the height transformed complex amplitude. In this study, we describe the effect of the amplitude factor of the height transformed complex amplitude in the FPC. The amplitude factor represents contrast information of the projected grating on the object, which corresponds to the reflectance property depending on 3D shape. We classify the amplitude factor by the 3D shape (flat surface, slope, edge, large step) and other factor (illumination condition, shadow, random pattern). We present the recognition characteristics with respect to the amplitude factor.

The ability of mobile robots to perceive and recognize environments is essential for autonomous navigation. To improve the performance of autonomous environment perception for mobile robots, it is important to effectively plan the next pose (position and orientation) of the sensor system at a current navigation state. In this paper, we propose a next-view-planning method for autonomous map construction needed for mobile robots with visual range sensor systems. The proposed view-planning method mimics the decision-making method of human beings, and uses the occlusion information reduced from the geometric relationship between the sensor view and objects as an important clue for the next sensor view planning. The proposed view-planning algorithms are developed in the following steps: 1) Given a prior map and range measurements sensed at a current location of the mobile robot, it is determined which parts in the map are interested in a view of solving the map uncertainty. 2) Based on the selected potential regions, some candidate poses of the sensor system for the next environment sensing are carefully generated. 3) The created candidates are evaluated by using a specially designed evaluation parameter, and the best one of them is selected as a next sensor position based on a fuzzy decision-making method. In this work, the principle of the view planning method is described in detail, and a series of experimental tests is performed to show the feasibility of the method for autonomous map building. For sensing the environments, an active trinocular vision sensor using laser structured light is utilized, which is mounted on the pan-tilt mechanism of the mobile robot, which is composed of a laser stripe projector and two cameras.

A measurement system that can measure the six-degree-of-freedom motions of arbitrary objects is proposed. The measurement system utilizes a special mirror looking like a triangular pyramid and having three reflective lateral surfaces, which is to be mounted on the objects of interest. Once a laser beam illuminates the top of the mirror, the mirror reflects and splits the beam into three beams. These reflected beams are detected by three position-sensitive detectors (PSD), respectively. Based on the signal outputs of the PSDs, the three-dimensional position and orientation of the mirror can be computed, which, in turn, determines the three dimensional position and orientation of the object. This paper proposes two options in applying the measurement principle, one of which has its laser beam source fixed in space, and the other controls the translational motion of the laser beam source to track the mirror. In this paper, the principle of the measurement is proved to be valid through theoretical analysis and experiments. And the advantages and disadvantages of both options are discussed.

Since at our laboratory, the basic configuration of the facial caricaturing system PICASSO has been constructed, it is strongly expected to get sufficient input image from a person who is naturally performing in front of the PICASSO camera system. From this viewpoint, we developed a face tracking PC system for capturing sufficient facial image especially in size by means of PTZ (Pan-Tilt-Zoom) camera collaborated with a fixed CCD camera. Irises are successfully recognized from the motion images captured from PTZ camera. These irises can be utilized to provide a key feature for realizing an automated facial recognizing system. In this system, a person performing naturally in poise and in facial expression within the scope of the fixed CCD camera can be stably tracked and the sufficient images in resolution of PTZ camera were successfully analyzed for iris recognition and facial parts extractions. This face tracking and face recognition system was characterized by a novel template replacement scheme among the successive image frames. Experimental results were also demonstated in this paper. This system works well in a practical speed 6-9fps on a usual PC connected to these cameras.

In this paper we present a novel method of object recognition utilizing a remote knowledge database for an autonomous robot. The developed robot has three robot arms with different sensors; two CCD cameras and haptic sensors. It can see, touch and move the target object from different directions. Referring to remote knowledge database of geometry and material, the robot observes and handles the objects to understand them including their physical characteristics.

We developed a Mobile Unit which purpose is to support memory retrieval of daily life. In this paper, we describe the two characteristic factors of this unit. (1)The behavior classification with an acceleration sensor. (2)Extracting the difference of environment with image processing technology. In (1), By analyzing power and frequency of an acceleration sensor which turns to gravity direction, the one's activities can be classified using some techniques to walk, stay, and so on. In (2), By extracting the difference between the beginning scene and the ending scene of a stay scene with image processing, the result which is done by user is recognized as the difference of environment. Using those 2 techniques, specific scenes of daily life can be extracted, and important information at the change of scenes can be realized to record. Especially we describe the effect to support retrieving important things, such as a thing left behind and a state of working halfway.

When inspection equipment based on visual recognition is installed in production sites, the recognition criterion is custom designed to satisfy the inspection specification, normally requiring constant attention and software modifications by specialty engineers. This paper proposes the Programless (Programming-free) visual recognition technology aiming at an automatic visual inspection system that has a self tuning mechanism and can also be easily tuned by factory workers. A camera system for visual inspection is mounted on a robot arm with flexible electronic components, which enables the users to perform positioning of the inspection equipment in an intuitive manner. The proposed system allows to provide a novel method of visual inspection and to reduce the technical complexity of the handling of visual inspection equipment.

3-D measurement of electronic components has been required to inspect the assembled PCB, visually. In many cases, the methods such as structure light methods were introduced into the practical inspection lines. The visual inspection systems with structure method are usually expensive and functionally limited. To solve the problem, a relative stereo method has been developed to the PCB visual inspection system. The relative stereo system has the unique feature of calibration free precise 3-D measurement. In the paper, the optical structure of the system and its processing method has been introduced.

In this study, a new display device based on physical contact and evanescent coupling is developed. The device consists of a glass substrate as a waveguide and electrostatically-driven multi-cantilevers on a silicon substrate. The incident light guided into the waveguide is transferred into the cantilevers due to the physical contact and the evanescent coupling between the waveguide and the multi-cantilevers when a voltage is applied. By the contact, the switched light is emitted from the side-edge of the contacted multi-cantilevers. The widely emitted output-light is realized by the integration of the cantilevers and the simplified structure. Therefore, the device is expected as the display application. The device is fabricated using MEMS process and bonding process. From the experimental results, a contrast of 0.9 is obtained at 170V, and the stable drive is attained up to 1KhZ. As a result, the usefulness of the device structure is confirmed.

Near-field optical storage using cantilever aperture tip is a promising way for next generation optical data storage. To enhance the speed of reading and writing data, gap between tip and media should be controlled fast and precisely within near field region. In this paper, several PZT actuators are analyzed for constructing dual servo control algorithm: coarse actuators (stack PZT, bimorph PZT) for media surface inclination and fine actuator (film PZT) for media surface roughness. Dynamic analysis of stack PZT, bimorph PZT, and film PZT are performed through the frequency response. Based on the frequency response and mathematical model, fast analog controller is designed and experimented. From experimental results, analog control of film PZT is 50 times faster than conventional stage. We get the 0.1 msec over 800nm step.

An optical actuator has some interesting characteristics, such as no generation of magnetic noise and receiving the energy remotely. A novel micromanipulator using photothermal effect is proposed. It consists of three optical fibers. One end of the fiber is cut for a bevel and painted with black color. A photothermal effect is occurred responding to the incident beam at the end of optical fiber. This effect makes the end of fiber stretching vibration. It can release a ball with 2.5 mm of diameter and a microlens currently held by light energy. The mechanism of photothermal vibration is also studied.

Conventional translation micromirrors for adaptive optics use attractive electrostatic force and therefore have two limitations: 1) the stroke is limited to less than one third of the initial gap distance between the mirror plate and the substrate. Normally the stroke is in the range of submicrometers; 2) stiction happens during operation. A novel translation micromirror, which uses a repulsive electrostatic force, is presented in this paper. This novel translation micromirror completely overcomes the limitations associated with conventional translation micromirrors and its stroke is not limited by the initial gap distance between the mirror plate and the substrate and therefore is able to achieve a much larger vertical stroke to modulate lights over a wider spectrum than that achieved by conventional translation micromirrors. The novel translation micromirror has no stiction problem and is highly compatible with mature surface micromachining technology. An analytical model is derived for the novel translation micromirror and prototypes are fabricated. The prototype of the novel translation micromirror, which is deliberately not optimized so it could be fabricated using MUMPS, achieved a vertical stroke of 1.75μm using a driving voltage of 50 volts, which is three times the stroke of conventional MUMPS translation micromirrors. It is expected that if standard surface micromachining is used instead of MUMPs, the design of the novel translation micromirror can be optimized and a much larger vertical stroke can be achieved.

The focus of this paper is to design a small incremental linear motor that could have the potential performance for light load applications. Due to its superior piezoelectric coefficients, PZN-PT (Pb(Zn_1/3Nb_2/3)O₃-PbTiO) is considered as a fungible material to conventional PZT. For the design of a lightweight linear motor, PZN-PT piezoelectric actuator was utilized to drive the motor due to its fast resopnse time and high piezoelectric coefficient. Furthermore, this paper presents the design and expected performances of a PZN-PT inchworm linear motor. We propose a new method of inchworm motion based on a novel electrode design using a single rectangular rod of PZN-PT single crystal piezoelectric material. The size of the monolithic PZN-PT linear motor can be very small (a length less than 5mm). The motor has a high-resolution positioning capability.

Optical controllers exploit lightwave technologies to implement different control strategies. It is possible to replace many of the electrical and mechanical components found in traditional linear and nonlinear controllers with optical analogues that increase the speed of signal processing or enhance system sensitivity. These optical sensors, switches, communication links, and actuators are largely immune from electromagnetic interference, exhibit low signal attenuation, provide secure flow of information, and are safe in hazardous or explosive environments. In addition, the energy in the light beam is one of the easiest forms of energy to shape and transmit through free space. The fundamental characteristics of several "control-by-light" systems are discussed in this paper. The proposed control system utilizes an acousto-optic deflector (AOD) to change the direction of the reshaped laser beam in response to the feedback error signal. The deflected beam strikes an array of photodetectors where each discrete detector represents a specific control action. One- and two-dimensional detector array configurations are explored for control. Although the controller designs can be implemented on optical breadboards using off-the-shelf optical devices, recent advances in nanotechnology would allow similar micro-scale optical controller to be fabricated at low cost.

The paper investigates optical implementation of feedback control of mechatronic actuators. The goal is to analyze experimentally the feasibility of optical closed loop control using optical components for signal transmission and control loop implementation. The analysis is carried out for position control of an electric actuator. Angular position and velocity are measured with optical sensors. Measurement signals are transmitted through a fiberglass to an optical controller. Position commands from the optical controller are transmitted through a fiber glass as analog optical signals to a photodetector. A power operational amplifier serves a driver for the electric motor. This feedback controller configuration is designed to have extensive optical implementation and permits to investigate the level of complexity of the control laws which can be implemented optically. Experimental results, obtained using a LabVIEW for data acquisition and processing, illustrate the advantages and the limitations of the proposed optical implementation of a feedback controller for mechatronic actuators.

In this paper, a small size monolithic XY scanner was designed and fabricated. The scanner has a flat triangular shape and consists of two 0.5mm-thick and 5mm-long lead-zinc-niobate-lead-titanate ((1x)Pb(ZnNb)O₃ - xPbTiO₃) or PZN-PT rods. The use of this material is critical to the reduction of the scanner size. The mechanical resonance characteristics for the PZN-PT rods and the assembled scanner were tested. The fabricated scanner provides a high resonance frequency and assured parallelism between the scanner and the sample surface. It enables a fast open loop control capability that naturally lends itself to use in scanning probe microscopy. The scanner and a self-sensing cantilever were integrated into a small-size atomic force microscope (AFM) design. A commercially available self-sensing cantilever and an additional actuator were used for contact scanning of a HOPG (Highly Ordered Pyrolytic Graphite) sample surface. The scanning performance of the scanner was verified by obtaining atomically resolved image of the HOPG surface.

Light activated optical circuits have several key advantages over conventional electronics because they are free from electrical current losses, resistive heat dissipation, and friction forces that greatly diminish system performance and efficiency. The effects of current leakage and power loss are also crucial design constaints in developing micro-electromechanical (MEMS) technology. An essential device for creating viable micro-optical circuitry is a robust photonic transistor that can act as a small signal switch and amplifier. The proposed photonic transistor is based on the complementary suppression-modulated light transmission properties of thin bacteriorhodopsin (bR) films. The light transmission properties exhibited by the thin film are controlled using the variable wavelength and intensity of the impinging light soruces. The light transmisison properties of the bR film are illustrated using a mathematical model for the two-state photoreaction system. The two-state model represents the longest lifetime in the bR photocycle, largest change in absorption maxima, and high photochemical stability. The optical response is proportional to changes in the light transmission properties of the biometrical, and therefore represents a viable material for creating optoelectronic devices.

Deformation measurement of a cantilever based on digital holographic interferometry is presented. The digital holograms of the cantilever before and after deformation are obtained by use of phase-shifting technique. By using the phase-shifting digital holography, the object wave information without zeroth order diffraction wave can be extracted. The phase difference of the reconstructed cantilever before and after deformation is obtained numerically using discrete Fresnel diffraction integral. Owing to the numerical analysis, the deformation of the cantilever can be obtained with high resolution of nanometer dimensions. The optical experimental results are shown to confirm our proposed method. Furthermore, we propose a mechanical movement free double-exposure method in phase-shifting digital holography using a laser diode.

The purpose of our study is to establish a robust communication, while keeping security and privacy, between a handheld communicator and the surrounding information environment. From the viewpoint of low power consumption, we have been developing a reflectivity modulating communication module composed of a liquid crystal light modulator and a corner-reflecting mirror sheet. We installed a corner-reflecting sheet instead of light scattering sheet in a handheld videogame machine with a display screen with a reflection-type liquid crystal. Infrared (IR) LED illuminator attached next to the IR camera of a base station illuminates all the room, and the terminal send their data to the base station by switching ON and OFF of the reflected IR beam. Intensity of reflected light differs with the position and the direction of the terminal, and sometimes the intensity of OFF signal at a certain condition is brighter than that of ON signal at another condition. To improve the communication quality, use of machine learning technique is a possibility of the solution. In this paper, we compare various machine learning techniques for the purpose of free space optical communication, and propose a new algorithm that improves the robustness of the data link. Evaluation using an actual free-space communication system is also described.

High speed remote radio control of systems containing multiple elements is usually performed using control channels for individual elements with multiple emitters and receivers at different conduits. If this signals are multiplexed and transmitted the bandwidth of each individual channel is reduced depending on the number of required channels. We propose a hybrid system that allows demultiplexing on flight time using a spatial light modulator to route the incoming signal into different out ports. The control signals are first multiplexed in frequency and a low speed routing control signal modulates it in amplitude. The combined signal is used to drive a diode laser. The proposed demultiplexing apparatus consists on a double 4f optical system in which the image of a dynamic liquid crystal display is imaged on the Fresnel plane. The control signal displays an amplitude binary mask on the dynamic spatial light modulator to route the signals to individual detectors located at the Fresnel plane in order to recover the individual channels. Afterward each channel is individually recovered. In order to demonstrate this system, multiple CATV signal were multiplexed with a digital control signal used to amplitude modulate the diode laser and tested on an optical bench. We also present the advantages of using such optical interconnects on the Fresnel plane over the Fraunhoffer plane.

A method for a beam shaping optics is proposed to convert a laser beam profile into required intensity distribution. A beam shaper with aspherical surface is calculated by Snell's law converted from measured irradiate beam profile to required profile. An acrylic resin is cut by a ball type grinding wheel consisted on diamond powder and polished by a cone type felt buff with alumina ceramics powder. A CNC lathe is used to control these tools position. A shaper with 4mm of thickness works to convert multi-mode beam profile with 13mm of diameter into a flat top cylinder with 1mm of diameter. Its working distance is 60mm. Its transmittance is over 98%.It suceed to generate uniform melting mark of 0.6mm diameter on the steel target by YAG laser process.

This paper describes a technique, based on the shape-from-focus method, to measure the shapes and heights of an array of solder bumps through visually inspecting an IC chip-size package (CSP). We devised a focus measure difference method for detecting the horizontal section of a bump quickly and precisely. This was where we selected bump image pixels that had the same focus measure at two specified focal points along the height axis. The selected pixels corresponded to the contour points of the section at the mid-height between the two focal points. We applied it to measuring the height of an array of 208 bumps that were 260 μm high arranged at intervals of 500 μm. By measuring the area of the horizontal section at a specified height and using it to fit a sphere in the bump, the height of the bump top was efficiently estimated to match the sphere top. This sphere fitting achieved a height measurement accuracy of 8 μm (2σ) at a resolution of 3.4 μm pixels on the bump using a 2/3-inch high resolution CCD camera with 1300 × 1030 pixels. The accuracy was sufficient to inspect the coplanarity of an array of solder bumps.

This paper describes a multispectral liquid drop analyzer for liquid chemical and physical properties analysis. Liquid drops formed at the tip of a liquid head are measured in parallel by a fiber sensor and a capacitive sensor. The fiber sensor works as follows: multispectral light sources are injected into the drop through an optical fiber and the total internal reflections and absorptions are detected by a photodetector. By combining fiber and capacitive sensor outputs, a drop speed independent one-dimensional waveform (liquid fingerprint) is generated. Liquid surface tension, refractive index and di-electric constant can be estimated from the fingerprint. To compare two fingerprints, the sensor outputs are normalized to have the same unit of measurement and drop starting position. After that, a reference liquid based calibration is applied to correct of fingerprint distortion due to variations in environment conditions, such as changes in temperature and humidity. Finally, a normalized correlation algorithm analyses the fingerprint difference. The repeatability and sensitivity of the system are demonstrated using different liquid samples. On-line applications show that the analyzer is able to detect 2% change in alcohol density.

The control of a robot manipulator by a human operator is often necessary in unstructured dynamic environments with unfamiliar objects. Remote teleoperation is required when human presence at the robot site is undesirable or difficult, such as in handling hazardous materials and operating in dangerous or inaccessible environments. Previous approaches have employed mechanical or other contacting interfaces which require unnatural motions for object manipulation tasks or hinder dexterous human motion. This paper presents a non-contacting method of teleoperating a robot manipulator by having the human operator perform the 3D human hand-arm motion that would naturally be used to compete an object manipulation task and tracking the motion with a stereo-camera system at a local site. The 3D human hand-arm motion is reconstructed at the remote robot site and is used to control the position and orientation of the robot manipulator end-effector in real-time. Images captured of the robot interacting with objects at the remote site provide visual feedback to the human operator. Tests in teleoperation of the robot manipulator have demonstrated the ability of the human to carry out object manipulator tasks remotely and the teleoperated robot manipulator system to copy human-arm motions in real-time.

At the Technical University Braunschweig a robot for the assembly of micro structures with an assembly positioning uncertainty below 1 μm is under development. In order to reach the demanded accuracy and robustness, an optical sensor is used. First a 2D-Sensor was tested. To reach the required cycle time a substantial reduction of the response time of this 2D-sensor was necessary. By the application of optimised 2D-algorithms the image processing takes place in video real time now. The extensive software update required an examination of the repeatability results of the 2D-sensor (standard deviation of 0.1 μm in both axes). An overview of the reproducibility and the dynamic cooperation with the parallel robot are presented. A prototype of a 3D-sensor for the robot was built. The reached reproducibility is presented. The 3D-sensor integrated into the robot allows the execution of assembly tasks with a positioning uncertainty better than 1 μm. For the assembly tasks the accurate position of the structures must be determined, which is marked by a defined pattern of circular marks. By photogrammetric measurement of the marks and a pattern recognition the spatial position of the structure is determined. The measurement volume is 7,5 x 11 x 6 mm³. To show the efficiency of the 3D-sensor a first test of its cooperation with the robot is described.

Innovative production processes and strategies from batch production to high volume scale are playing a decisive role in generating microsystems economically. In particular assembly processes are crucial operations during the production of microsystems. Due to large batch sizes many microsystems can be produced economically by conventional assembly techniques using specialized and highly automated assembly systems. At laboratory stage microsystems are mostly assembled by hand. Between these extremes there is a wide field of small and middle sized batch production wherefore common automated solutions rarely are profitable. For assembly processes at these batch sizes a flexible automated assembly system has been developed at the iwb. It is based on a modular design. Actuators like grippers, dispensers or other process tools can easily be attached due to a special tool changing system. Therefore new joining techniques can easily be implemented. A force-sensor and a vision system are integrated into the tool head. The automated assembly processes are based on different optical sensors and smart actuators like high-accuracy robots or linear-motors. A fiber optic sensor is integrated in the dispensing module to measure contactless the clearance between the dispense needle and the substrate. Robot vision systems using the strategy of optical pattern recognition are also implemented as modules. In combination with relative positioning strategies, an assembly accuracy of the assembly system of less than 3 μm can be realized. A laser system is used for manufacturing processes like soldering.

In this paper, we present an automation technique that yields high performance, low cost optoelectronic alignment and packaging through the use of intelligent control theory and system-level modeling. Our control loop design is based on model predictive control, previously popularized in process and other control industries. Our approach is to build an a priori knowledge model, specific to the assembled package’s optical power propagation characteristics, and use this to set the initial "feed-forward" conditions of the automation system. In addition to this feed-forward model, our controller is designed with feedback components, along with the inclusion of a built in optical power sensor. The optical modeling is performed with the rigorous scalar Rayleigh-Sommerfeld formulation, efficiently solved using an angular spectrum technique. One of the benefits of using our knowledge based control technique is that the efficiency of the automation process can be increased, as the number of alignment steps can be greatly reduced. An additional benefit of our technique is that it can reduce the possibility that attachment between optical components will occur at local power maximums, instead of the global maximum of the power distribution. Therefore, our technique improves system performance, while reducing the overall cost of the automation process.

This paper presents a visual servoing methodology to realize a vision-based chip mounting, which makes it possible to mount micro chips to PCB with high precision. Because of continued down-sizing of electronic products, density of electronic parts to be packed into PCBs is ever getting higher and higher, thus requiring surface mounting technologies (SMT) to be more precise and productive. Down-sizing also tends toward miniaturization of components. The increased package density and component miniaturization require not only delicate handling of small chips but also delicate placement of them on PCBs. In order to place the small chips on the pattern of PCB, highly accurate mounting process is ultimately required. This is because it is difficult to place the chips a specified position with tiny small contact force due to uncertainty involved with the kinematic calibration error of the mounter and the deformation of PCB. In this paper, to avoid this difficulty we propose a visual servoing method for the chip mounting system. In such a way, the proposed visual servoing technique guarantees that chips are accurately placed at the desired positions. To achieve this, a CCD camera views the chip mounted with the mark on PCB simultaneously, and then the edge features and corner features are extracted from the acquired image. The proposed visual servoing algorithm generates the motion of the mounter to align the features of the chip with the ones of the mark on PCB on-line. The result obtained from real time experiment is discussed from the view points of practicality of the proposed method.

Intrinsic properties of liquid crystals spatial light modulators, as sampling devices are commonly seen as disadvantages in opto-digital data processing; because the loss of light power due the replicated spectrum at the sampling process. We claim this apparent disadvantage can be seen as an advantage if we consider to filter Q replicates of the Sampled image Fourier spectrum. Moreover, we can use this technique to improve the signal to noise ratio (S/N), or to develop several operations in parallel over the same input data. In this work we are interestd in improving the S/N in opto-digital image substraction. Experimental results which reveal the advantage of this technique are shown.

An approach to improve the accuracy of phase-shift interferometry is presented. In this paper, an algorithm for estimating phase shift step errors is demonstrated. The algorithm is based on the fact that the sinusoidal intensity data from the same pixels of two interferograms with different phase shifts form an elliptic Lissajous curve. The elliptic Lissajous curve can be fitted by the least squares method from which the phase shift steps can be accurately estimated. The estimated phase shift step errors are then compensated to measure 3D topography of specimen. In addition, the approach provides a simple technique of measuring 3D topography without sophisticated actuation mechanism. Simulations and experiments also demonstrate that the intensity noise in interferograms provides very small effect on the accuracy of the algorithm.

We propose a new vertical-scanning profilometry which has high-speed in measurements. The proposed profilometry measures 3-D shape by using phase-shifting technique with large phase shifts around 6 π ± π /2. With such large shifts, 2 π phase ambiguities normally suffer precise measurements of shapes. Therefore, the profilometry is equipped with two short-coherent-light sources of different mean-wavelength and alternately acquires interferogram for each wavelength every large phase shift (one interferogram /one shift). From the interferograms acquired through vertical scanning , it searches the highest-contrast frame position regarding the each wavelength and calculates the phases with the searched frames. From the phases of the two wavelengths and a vertical-scanning step height (relating to the large phase shifting), 3-D profile is calculated with a precision as high as a well-known phase-shifting interferometry.

This paper presents a new scanning mirror structure. Large area (mm-order) scanning mirrors have been studied and developed due to the many applications where mm-order light beam size is present like optical microscopes and instrumentation systems. In the proposed structure the actuation and reflection mechanisms were separated in order to provide a more flexible and accurate design that considers the specific needs of each one. The new structure consists of two square rotors linked to a fixed frame by two torsion bars, a third torsion bar connect both rotors. The electromagnetic induction actuated scanners were made using bulk silicon micromachining technology, thin film techniques and mechancial assembly. The maximum optical deflection angle was 8.0°pp at the first resonant frequency of 1316Hz with a quality factor of Q=200. The second resonant frequency was 2542Hz with optical angle of 6°pp and a quality factor of Q=422.

This paper detailedly presents a moving robot with multi-sensor, which can be programmed on-line. The main controller of moving robot is designed based on the controlling function of 80C552 microprocessor, and the graph programming software developed with VC++ 6.0 is used, making the robot easy to be manipulated. By downloading programs to its controller through the serial port of computers, the moving robot can avoid obstructions, recognize sound and move off-line automatically depending on various sensors and micro operating system.

Automatic deephole admeasuring apparatus is an automatic synthetical multiparameter instrument which makes use of the automatic crawl capability of pipeline robot to drive the measuring device moving forward and adopts optical measuring method to measure deephole inside diameter, interior surface roughness concentration and straightness accuracy. This apparatus has distinguishability of 1 μm with measuring precision less than 5 μm, range of roughness between 1.6 and 2.0, measuring error between 1/8 to 1/4 grade; straightness accuracy measuring precision of deephole as long as 10 meters is less than 100 μm.