A large format backside illuminated CCD has been developed by THOMSON TCS. This thinned CCD is mainly dedicated to scientific imaging field where high sensitivity and low noise operation are required. This product was designed specifically for the Very Large Telescope optical instruments needs, and supported by ESO organization. The imaging area is a 2 k X 2 k 15 micrometers square pixel array, with 100% optical aperture, arranged in a full frame organization. Charge transfer operates in the four phase mode, and thus provide large handling capability in MPP mode. This device offers two low noise output ports that reach typically 4 e- rms noise in slow scan operation (at 50 kHz typical data rate) at -40 degree(s)C. Thanks to THOMSON 1.5 micrometers technology, and to its wafer thinning technology, this device features a very high quantum efficiency over a wide spectrum range (typically 80% peak quantum efficiency). Its specific thinning and assembly processes make it suitable for construction of 2 X N CCD mosaic (three sides are buttable), with less than 400 micrometers dead gap between sensitive areas and good flatness performance (typically 15 micrometers peak to peak is achieved). Electro-optical characterization results are described in this paper as well as the specific assembly process and butting technique.

Charge coupled devices have been the dominant solid state detector array in the visible due to their relatively simple design and easy implementation. With recent advances in lithographic techniques, arrays having smaller photosite dimensions and an increased number of pixels have become available. Further advances in large format CCDs have been limited by charge transfer efficiency (CTE) of photoelectrons to the readout amplifier. The increased number of pixel transfers in large arrays can degrade image quality and MTF unless even higher CTEs are achieved. Multiplexer designs that remove the need for thousands of charge transfers can bypass these CTE limitations. One such focal plane architecture is the CID or charge injection device. This paper presents results obtained with one particular CID based system. The array is housed in a dewar capable of liquid nitrogen operation. The output signal from the array is amplified with a nearby low noise preamplifier before digitization. Results on injection efficiency, readout noise, and other pertinent CID parameters, are presented obtained from this device preamplifier as well as specific experiments.

This paper will describe in some detail a new large area CCD image sensor designed specifically to be used either as a single imager or assembled in large, tightly configured mosaics of CCDs. The device has 2048 X 4096, 15 micrometers pixels. Performance data are presented on both front- and back-illuminated parts.

We have developed a 6032 element, 32 stage Tri-linear Time Delay and Integration Focal Plane Array for high resolution color imaging applications. The sensor offers an improvement of a factor of 10 over comparable line scan CCD sensors. The imager architecture utilizes three individual TDI arrays, with a new multi-layer dielectric interference film color filter that helps achieve accurate spectral separation in the three primary color bands centered at 450 nm, 550 nm and 650 nm respectively. We have also incorporated in the sensor a new programmable responsivity mechanism. This is achieved by an on-chip imaging area size selection mechanism. The sensors operate in the time-delay-and-integration mode. Depending on the level of illumination the user is able to dynamically select the number of TDI stages per sensor. The total number of TDI gain stages can be selected in blocks of 32, 16, 8, or 4. This property provides improvement in the dynamic range of the sensor by extending the response of the CCD over a wide range of illumination levels. Depending on the light energy incident on the sensor, the user can dynamically vary the number of TDI stages used for integrating an incident scene. The result is a programmable responsivity depending on the level of incident illumination. Such a technique optimizes the performance of the CCD color sensor. Given the resolution, speed, sensitivity and the dynamic range, this sensor is suitable for a variety of color imaging applications including multi-media, document scanning, pre-press scanning, medical and machine vision.

Benzocyclobutene (BCB) is a commercial product of The Dow Chemical Company sold under the irade name Cyclocene TM Electronic Resins for use in multi-chip modules and interlayer dielectric applications. New versions of the material are being developed as both overcoat and planarization materials for use in active and passive components of flat panel displays and charge coupled devices. Specifically, BCB has shown excellent results for color filter elements as either an overcoat or undercoat material.

A JET-X (Joint European X-ray Telescope) CCD fabricated on a high-purity (1.5 k(Omega) cm), 65 micrometers thick epi-layer, on a 550 micrometers thick p⁺ substrate has been developed for X-ray astronomy. A 3D numerical analysis for evaluating a charge handling capability and charge transfer efficiency of a JET-X CCD using a static and transient simulation has been performed. A supplementary channel technique is analyzed by the 3D simulation. A static maximum charge capacity was found to be 60040 electrons under a full- well condition. The effect of an output gate voltage on charge transfer between the last well and an output diffusion node was observed and an optimum output gate voltage for efficiency charge transfer was found to be 3 or 4 V. A time-dependent simulation was performed to observe the dark current contribution, to very the static, full-well capacity and to estimate the charge transfer efficiency (CTE). The dark current source has found to be < 1 electron per pixel for a clock cycle, T equals 1.85 microsecond(s) . The CTE was > 99.999% for a pixel clock cycle of 4.2 ns with a fall time of 0.4 ns. The dynamic, full-well capacity was higher than the static, full-well capacity by appr. 1.8%. The 3D simulated result showed a higher charge capability by appr. 18% than result from a 1D model.

Design considerations and description of a CMOS Active Pixel image sensor (APS) star tracker are reported. APS technology has been thought of being most appropriate for guidance and navigation. However, making APS useful for future star tracker missions means a few challenges have to be overcome. A wider dynamic range is required, while fill factor ought to be high and simple geometry of the pixel active area is desired. These requirements are analyzed and tradeoff considerations are explained for a practical future celestial tracker. A 64 X 64 element CMOS APS array with individual pixel reset is described.

A novel CMOS active pixel sensor structure has been designed, fabricated and characterized. It greatly increases the working range for all imaging applications in which the optical signal information to be detected is superimposed on a large DC offset signal. This is achieved by subtracting an offset current at each pixel's photosite. The offset current can be programmed individually by an external programming voltage. Experimental results from a single pixel test-cell fabricated on a standard 2 (mu) CMOS-process show a programmable offset signal range of > 150 dB with a dynamic range of > 60 dB of the photo detector itself. To achieve a similar performance using conventional imaging techniques would require an imager with a dynamic range of > 150 dB.

Previously realized circuits for solid state light sensitive arrays fabricated in thin-film technology show a comparatively small dynamic range. A wide dynamic range as well as a large sensitivity of the output signal to the change of the input signal (irradiance) is required. A circuit which satisfies these requirements is described. The circuit consists of two thin-film transistors (TFTs), a photoconductor, and a fixed resistor. The photoconductor is used to detect the incoming light whereas the first TFT generates the logarithm of the photocurrent in form of a voltage. The second TFT and the fixed resistor invert the signal of the first stage and determine the slope of the output signal. The sensor's characteristic can be shifted over a wide dynamic range by using an appropriate control voltage. The circuit is simulated by SPICE, employing a special model for the TFTs. The sensor's characteristics show a large dynamic range of 4 decades when shifted by means of the control voltage. At the same time it exhibits a high sensitivity of 1.8 V/W/m² due to the steep slope of its characteristics. This circuit was then built in a-Si:H thin-film technology. The measured characteristics match very well the simulations.

Two CMOS image sensor concepts, developed for motion extraction, are proposed. The algorithm implemented in each pixel is either: the calculation of the temporal variation of the difference of the logarithm of intensity in two adjacent pixels; or a more general implementation of the spatial and temporal filtering over the local neighborhood. Temporal differencing yields peak in the response of pixels with changing intensity. The spatial differencing provides high-pass filtering and invariance to time-varying external lighting. We also compare two ways to use this sensor module to compute the velocity of edges moving along the sensor. In one implementation, the sensor is used as an input for a correlation algorithm, calculating the optical flow vector. The other possibility is to detect motion locally in each pixel, and measuring the time of switching between adjacent pixels which detect the motion.

We describe a compact, vacuum compatible, large format, charge-coupled device (CCD) camera for scientific imaging and detection of 1 - 100 keV x-rays in experiments at the Lawrence Livermore National Laboratory JANUS-1 ps laser. A standard, front-illuminated, multi-pin phase device with 250 k electron full well capacity, low dark current (10 pA/cm² at 20 degree(s)C) and low read noise (5 electrons rms) is cooled to -35 degree(s)C to give the camera excellent 15-bit dynamic range and signal-to-noise response. The intensity and x-ray energy linear response have been determined for optical and x-ray (< 65 keV) photons and are found to be in excellent agreement. Departure from linearity has been measured to be less than 0.7%. The inherent linearity and energy dispersive characteristics of CCD cameras are well suited for hard x-ray photon counting techniques in scientific applications. X-rays absorbed within the depletion and field-free regions can be distinguished by studying the pulse height spectrum. Results are presented for the detection of 1 - 100 keV Bremsstrahlung continuum, K-shell and L-shell fluorescence spectra emitted from high intensity (10¹⁸ W cm^-2), 500 fs laser-produced plasmas.

The A62Se3 - Si02 - Si structure with A62Ses for writing a.nd Si for readout :h.a.s been suggested ior Xra. y im.a.ge a.pplica.tion. In this structure th.e recording ha.ve been. ca.:rri.ed out under sim:ulta.n.eous projectioit oi a.n X-ray pa.ttern on the top semitransparent contact a.nd a.pplica.tion of the external volta.ge, but the erasing - under illumin.a.tion. of light from the ra.n.ge of th.e .4s2S es funda.men.tal a.bsorption. a.n.d a.pplica.tion. of th.e opposite pola.rity external voltage. Th.e cha.."ging kinetics was revealed to be described by exponential la.w with. saturation.. It was found th.at th.e charge is a.ccumu.la.ted at th.e deep tra.ps disposed a.t th.e As2Se3- Si02 interface. Th.is structure ma.ke it possible to work m the integration small signa.i regime but the space separation. of recording and read.out layers provides un.destromg repetition.al readout of image. It was established th.at th.e da.rk rela.xa.tion of a.ccumula.ted cha.rge ta.kes place due to th.e.rmo-:iield emission of holes from t:ra.ps according to Pool-Frenkel la.w. Keywords: solid state detector, image device structure, X-ray, area meter, amorph.ous semiconductor

We have investigated main characteristics of linear multielement infrared receiver (sensitivity, time constant) and their dependencies from geometrical and thermophysical parameters. For this purpose 3D mathematical model of heat diffusion processes in the receiver has been constructed. Obtained numerical solution gives us wide possibilities to estimate theoretically main properties of exploring system.

This paper describes the technology used in a new generation of digital cameras. The cameras all use full-frame image sensors optimized for producing still images in electronic cameras. The color CCDs incorporate the Bayer color filter array pattern, lateral overflow drain antiblooming protection, accumulation mode timing, and progressive scan readout. The three cameras described have CCDs with 6.3 million, 1.6 million, and 400 thousand pixels, with 9 micron square pixels and a 3:2 image aspect ratio. They use a firmware-based digital camera architecture to maximize flexibility and image quality. Extensive digital image processing is performed in the host computer, as the images are downloaded from the camera. This enables the cameras to use sophisticated image processing algorithms that can be easily upgraded in the field and customized for special customer applications.

A PC digital still camera based on a mega-pixel CCD area array is presented. This camera can be used as an image input device in the applications of multimedia. The system is equipped with a conventional camera lens, and an 1.5 mega-pixel progressive color CCD area array. A driving circuit is designed to control the focusing and the iris of the lens in order to perform auto/manual focus and auto/manual exposure functions. The CCD image sensor is also used as a sensor for auto focus, auto exposure and auto white balance control. This could eliminate the additional sensors. A shutter wheel is built for continuous exposure, which allows the host computer to display the entire field of view for preview without using an additional optical viewfinder. The digital signal processing is implemented in a computer software for interpolation, color reproduction, and edge enhancement, etc. A SCSI interface is used to connect to a host computer for data and command transmission.

We describe in this paper an evaluation system for completely characterizing a digital still camera developed at our laboratory. The CCD sensor array of the camera was first used to assess the modulation transfer function of the lens system, the calibrated lens was then used to measure the single-chip color CCD sensor array. Accordingly, performance of the complete camera system can be evaluated as a whole. Measurements performed on this imaging system include dynamic range, signal to noise ratio, modulation transfer function and contrast transfer function of the lens as well as the CCD. With the help of a standard color test plate and prior knowledge about the spectral distributions, spectral response of the color filters coated on the CCD were computed approximately. The results of our experiments were compared with the data sheet provided and the theoretical calculation. The apparatuses needed for measuring the entire imaging system include only calibrated light source and standard test patterns, while the data access subsystem was part of the digital camera and the mathematical analyzing tools are developed on a personal computer. Consequently, system design was simplified and the evaluation process was made more efficient and cost effective.

Key progress in the development of non-TV format CCD cameras has been made in high- speed scanning rate applications. VGA format has become an industrial standard for fast speed machine vision applications. While it is important to operate a camera in high speed applications with asynchronous reset, electronic shutter, or variable integration functions in order to capture dynamic motion, conventional strobe light applications are also still important. A combinatorial-scan CCD provides VGA and high speed scanning as well as frame mode interlace scanning for such applications. The main issues of this study are: (1) the structure of combinatorial-scan CCD's, (2) a fast dump structure to operate partial scan functions and asynchronous read-out control, (3) normal interlace operation with frame mode for strobe light applications, and (4) digital processing for dual channel or special functions to achieve fast frame mode. The first model was developed for the VGA format with a single- channel CCD output requiring scanning twice that of a normal CCD. The second model studies the combination of progressive and interlace scan functions. The third model has built- in digital processing and frame memory to capture the dual channel outputs of the CCD and scan-convert to various formats.

We describe a series of five CCD cameras designed by Gordian for low light-level applications. The first device is a low-cost non-imaging astronomical autoguiding tracker based on the Texas Instruments TC255 CCD chip and an MC6811 microcontroller. Mounting off-axis, it provides standardized tracking-motor signals for any telescope with a dual-axis drive corrector, automatically compensating for the mechanical peculiarities of the drive, set- up factors, and pointing errors. The tracker can guide to +/- 1 arcsec on an 8th magnitude star when used with an 8' aperture, f/10 telescope. The basic autoguider design has been extended to produce self-contained 8-bit and 16-bit imaging cameras with autoguiding functionality. Images are buffered in PSRAM, then relayed to a host PC via an RS-232 serial connection. The addition of regulated thermoelectric cooling reduces CCD thermal noise and alleviates dark current saturation. Gordian has also designed two high-resolution cameras based on the Kodak KAF-0400 and KAF-1600 CCDs. The cameras produce 16-bit images with 768 X 512 pixels or 1536 X 1024 pixels, respectively. Pixel size is 9 micrometers square. The camera head contains the CCD, thermoelectric cooling mechanism, analog electronics, and a custom-designed electromechanical shutter based on Flexinol^TM actuator wire. A separate base unit houses a Motorola 68306 microprocessor and associated electronics for telescope control and on-board image processing. A stepper-motor based filter wheel can be attached directly to the camera head. The camera communicates with a personal computer via SCSI or serial connection. Software for the host PC provides additional control options, data storage, and image processing capability.

Camcorders record motion video while electronic still cameras and computer cameras capture still images. Most camcorders are NTSC resolution, analog-based systems. A megapixel progressive scan motion and photographic still system is described in this paper which uses a high-resolution color CCD imager and the analog signal processing in Kodak Digital Science^TM KASP series integrated circuits. For maintaining the 30 frames per second (fps) the sensor is clocked out at 37.5 MHz. The video output from a CCD sensor is in the shape of a high-frequency digital clock whose reset pedestal is embedded with KTC noise. Hence a correlated double sample and hold (CDS/H) circuit can give the difference between the reset pedestal and the video for every pixel and thus eliminate the KTC noise and 1/f noise. Every camera using a CCD sensor will have a CDS/H. Most cameras implement CDS/H at 40 MHz in the form of a filter or use general purpose track and hold amplifiers. Both options are bulky, expensive, and power hungry. Kodak has developed a monolithic 40 MHz analog signal processor (KASP-140G) with CDS/H using QuickTile standard cell architecture of Tektronix, Inc., using the Quickic tool set and the TSPICE analog circuit modeling tools. The paper describes the high-resolution progressive scan sensor, data rate speed issues, and present implementation methodologies of CDS/H at 40 MHz and compares the power, performance and cost savings by using KASP-140G. Kodak has designed four different analog signal processing chips with varying levels of power consumption and speed for analog signal processing of video output of CCD sensors. The analog application specific integrated circuits have been characterized for noise, clock feedthrough, acquisition time, linearity, variable gain, line rate clamp, black muxing, affect of temperature variation on chip performance, and droop. The ASP chips have met their design specifications.

The transition between electronic and chemical photography will require the digitization of large volumes of source material currently on photographic film. For efficient conversion, the resolution and number of quantization levels must be well matched to the information content of the film and the display device. The grain noise levels for 5 readily available commercial color photographic films were measured. These measurements are compared to an empirical model that predicts the noise in the transmission coefficient from the manufacturers' quoted granularity values. For all the emulsions measured, the film grain noise does not seem to warrant retaining more than 8 bits, i.e. 256 levels, in the green channel for images displayed on a conventional CRT. The noise in the red and blue channels was found to be equal to or greater than that measured in the green channel with the behavior being highly dependent on the specific emulsion type. For monochrome photographic emulsions, the grain noise is usually assumed to vary as the inverse of the square root of the sampling area. For the five color emulsions measured, the noise was found to vary as the inverse cube root of the sampling area or slower. These results suggest that it can be difficult to significantly decrease image noise by increasing the film format size.

Astronomical CCD controllers are being called upon to operate a wide variety of CCDs in a range of ground-based astronomical applications. These include operation of several CCDs in the same focal plane (mosaics), simultaneous readout from two or four corners of the same CCD (multiple readout), readout of only a small region or number of regions of a single CCD (sub-image or region of interest readout), continuous readout of devices for drift scan observations, differential imaging for low contrast polarimetric or spectroscopic observations and very fast readout of small devices for wavefront sensing in adaptive optics systems. These applications all require that the controller electronics not contribute significantly to the readout noise of the CCD, that the dynamic range of the CCD by fully sampled (except for wavefront sensors), that the CCD be read out as quickly as possible from one or more readout ports, and that considerably flexibility in readout modes (binning, skipping and signal sampling) and device format exist. A further requirement imposed by some institutions is that a single controller design be used for all their CCD instruments to minimize maintenance and development efforts. A controller design recently upgraded to meet these requirements is reviewed. It uses a sequencer built with a programmable DSP to provide user flexibility combined with fast 16-bit A/D converters on a programmable video processor chain to provide either fast or slow readouts.

A new CCD based field acquisition and telescope guiding camera is being designed and built at UCO/Lick Observatory. Our goal is a camera which is fully computer controllable, compact in size, versatile enough to provide a wide variety of image acquisition modes, and able to operate with a wide variety of CCD detectors. The camera will improve our remote-observing capabilities since it will be easy to control the camera and obtain images over the Observatory computer network. To achieve the desired level of operating flexibility, the design incorporates state-of-the-art technologies such as high density, high speed programmable logic devices and non-volatile static memory. Various types of CCDs can be used in this system without major modification of the hardware or software. Though fully computer controllable, the camera can be operated as a stand-alone unit with most operating parameters set locally. A stand-alone display subsystem is also available. A thermoelectric device is used to cool the CCD to about -45c. Integration times can be varied over a range of 0.1 to 1000 seconds. High speed pixel skipping in both horizontal and vertical directions allows us to quickly access a selected subarea of the detector. Three different read out speeds allow the astronomer to select between high-speed/high-noise and low-speed/low-noise operation. On- chip pixel binning and MPP operation are also selectable options. This system can provide automatic sky level measurement and subtraction to accommodate dynamically changing background levels.

The conversion of the Multiple Mirror Telescope from six 1.8 m primary mirrors to a single 6.5 m primary will significantly increase its capability for imaging. The f/5 configuration will provide a corrected field of view for imaging that is flat and 30 arcminutes in diameter. The image quality in the absence of atmospheric seeing is 0'.1 over the full field. We are currently designing a camera system to take advantage of this large field. The proposed direct imaging system will be located at the Cassegrain focus of the telescope, behind a three-element refractive corrector. We will use an array of 8 X 4 three-edge-buttable CCDs, each with 2048 X 4096 pixels and two output amplifiers. This will provide a field of view of 24' X 24'. With a new packaging scheme we will reduce the gap along the readout edge to a few millimeters. The pixel size is 15 microns, or 0'.09, well sampling the point-spread- function. In many applications it will be possible to bin the pixels, thus reducing the amount of data (500 Mb per read at full resolution). The back-illuminated CCDs will be thinned and anti- reflection coated to provide high quantum efficiency from 320 to 1000 nm. The camera system will be useful for many studies requiring both a large collecting area and large area coverage on the sky. Planned projects include redshift and photometric surveys of faint galaxies, searches for high-redshift quasars and searches for objects in the outer solar system.

At the University of Hawaii, Institute for Astronomy, we are carrying out several activities to develop CCDs for astronomical observation. In this paper, we report the test results of devices from three CCD manufacturers: Loral, Hamamatsu Photonics and MIT/Lincoln Laboratory.

In the last years, the Charge Coupled Device (CCD) detectors have had a great development: 2048 X 2048 pixel formats are routinely produced by silicon foundries with good electro- optical characteristics. Scientific CCDs now, not only offer the ability to be read from more than one output, but they can also be buttable to form mosaics in order to cover a larger field of view, requirement posed by the current telescope technology. The Italian National Telescope GALILEO (TNG) will support a large set of visual and near IR detectors dedicated to scientific measurements at the focal plane. Also tracking systems and Shack-Hartmann wavefront analyzers will be based on CCD technology. Due to the number of camera systems to be routinely operated, the possibility to have uniformed interaction and configuration of systems is emerged as an important requirement for this crucial part of the telescope. In this paper the detector and instrument plan foreseen for the TNG telescope will be presented on the first part, while on the second we will present the CCD controller, now at the end of development. Here presented is a modular system based on digital signal processors and transputer modules. It is interfaced to host computers (PCs, workstations or VME crates) via optical fibers and a specially developed VME-VSB interface board.

Traditionally, scientific camera systems were partitioned with a `camera head' containing the CCD and its support circuitry and a camera controller, which provided analog to digital conversion, timing, control, computer interfacing, and power. A new, unitized high performance scientific CCD camera with dual speed readout at 1 X 10⁶ or 5 X 10⁶ pixels per second, 12 bit digital gray scale, high performance thermoelectric cooling, and built in composite video output is described. This camera provides all digital, analog, and cooling functions in a single compact unit. The new system incorporates the A/C converter, timing, control and computer interfacing in the camera, with the power supply remaining a separate remote unit. A 100 Mbyte/second serial link transfers data over copper or fiber media to a variety of host computers, including Sun, SGI, SCSI, PCI, EISA, and Apple Macintosh. Having all the digital and analog functions in the camera made it possible to modify this system for the Woods Hole Oceanographic Institution for use on a remote controlled submersible vehicle. The oceanographic version achieves 16 bit dynamic range at 1.5 X 10⁵ pixels/second, can be operated at depths of 3 kilometers, and transfers data to the surface via a real time fiber optic link.

Characterization of CCDs is extremely important when developing scientific detectors. If CCD foundries are used to produce the devices, the foundries require feedback to maintain a quality process. In this case, the users require fairly automated testing to evaluate the large number of devices obtained from even a single lot run. We have developed a CCD characterization facility which is used to evaluate these foundry devices as well as commercial scientific images. Our test capabilities include automated QE measurements, X-ray CTE and gain calibration, optical illumination from 200 nm - 1200 nm, and dark current and read noise characterization. We can also make interferometric flatness measurements of the devices. A cryogenic probe station for wafer testing is being developed to extend some of these tests to the wafer level. We discuss in this paper our facilities and techniques to measure the quantum efficiency (QE) of scientific CCDs. QE (along with read noise) is perhaps the most important parameter for many classes of astronomical observations when working at very low light levels. It is also the most useful parameter for evaluating the quality of backside processing when developing back illuminated CCDs.

An ultraviolet-sensitive scientific CCD camera has been tested at a power reactor facility to image the faint Cerenkov light from irradiated nuclear fuel. The instrument mates custom optical components (lens, UV-pass filter) to a commercial scientific camera (Astrocam 4100) with a coated frame-transfer CCD chip (EEV 37-10) to produce 12-bit images of 512 X 512 pixels at a near-real-time frame rate. A 250-mm f/2.6 catadioptric lens has been designed with transmissive optics optimized for this application, incorporating color correction for viewing through 10 m of water. The filter has an average transmission of 80% from 280 to 320 nm, with visible-light transmission of less than 0.01% to block artificial lighting in the fuel bay. Measurements were made with this instrument at the Ringhals Nuclear Power Plant, Varobacka, Sweden. Both fuel and non-fuel assemblies of boiling-water reactor type were studied. Performance is superior to that of the earlier Cerenkov viewing devices based on image intensifier tubes. Increased sensitivity extends the range of the Cerenkov verification technique to fuel with older discharge dates. Increased resolution allows fine details of the fuel to be examined for higher-confidence safeguards verification. Sample digital images are presented, and the advantages to irradiated-fuel verification of image quantitation, storage, transmission, and processing are discussed.

Power lasers are more and more used in aerospace industry or automobile industry; their widespread use through different processes such as: welding, drilling or coating, in order to perform some surface treatments of material, requires a better understanding. In order to control the quality of the process, many technics have been developed, but most of them are based on a post-mortem analysis of the samples, and/or require an important financial investment. Welding, coating or other material treatments involving material transformations are often controlled with a metallurgical analysis. We here propose a new method, a new approach of the phenomena, we control the industrial process during the application. For this, we use information provided by two CCD cameras. One supplies information related to the intensity, and geometry of the melted surface, the second about the shape of the powder distribution within the laser beam. We use data provided by post-mortem metallurgical analysis and correlate those informations with parameters measured by both CCD, we create a datas bank which represents the relation between the measured parameters and the quality of the coating. Both informations, provided by the 2 CCD cameras allows us to optimize the industrial process. We are actually working on the real time aspect of the application and expect an implementation of the system.

A photon counting system, utilizing a CID as the imager sensor, is under development at RIT. The system integrates a programmable, DSP based driver system capable of generating fast readout and sub-array dynamic control sequences. A high speed event recognition and centroiding computation task is performed by a dedicated board, based on field programmable gate array technology, which provides the necessary spatial resolution while maintaining high data throughput capability. The system architecture is flexible and capable of handling different CID array architectures and sizes. Preliminary performance results are presented, and characteristics of CIDs, such as subarray injection, that impact the total possible throughput, and therefore, the dynamic range, are discussed.

Scanner noise is one of the fundamental parameters of image quality. In this paper, we present an algorithm developed to derive the noise of a scanner using the 2D Wiener spectra of the test pattern and the scanner's MTF. The Wiener spectra of the test pattern was measured and its contribution to the measured RMS noise was estimated by integrating the volume under the product of the test pattern Wiener spectra and the scanner's MTF. The test pattern contribution was then removed from the measured noise. The derived noise agrees very well with the noise model for both drum scanner and CCD scanners. The structured 1D noise is also of interest especially when evaluating CCD scanner systems. A method was described to accurately determine 1D structured noise by averaging over fast scan and slow scan directions. Finally an experiment was conducted to verify the noise measurement technique. The true noise of a drum scanner was measured at its analog output terminal, and was compared to the noise estimated with the proposed new noise metric. The agreement between hardware measured noise and the estimated noise is very good with RMS error of less than 0.001 in reflectance unit. With this new technique, we can effectively improve the noise measurement accuracy by a factor of up to 500% for a photographic test pattern.

NHK led the world in developing a high-sensitivity Super-HARP pickup tube using the avalanche multiplication effect. The authors have now developed an improved version (2/3- inch, with electromagnetic focusing and electromagnetic deflection) which is eight times more sensitive and has much better lag characteristics. The handy New Super-HARP color camera which uses this newly-developed pickup tube has higher sensitivity than human vision (2000 lux, F/110 equivalent), negligible low lag, and a limiting resolution of over 700 TV lines. It will be a powerful tool in emergency news gathering at night, the production of scientific programs, and other applications.

Nine imagers that exploit distinctive CID properties and incorporate on-chip amplifier configurations (including preamplifier/pixel) were developed for use in automation, nuclear and scientific applications. TV compatible (11 mm) formats of 768_H X 575_V (European) and 755_H X 484_V (domestic-RS170) were fabricated for radiation- hardened product cameras. Operating CIDs provided excellent signal-to-noise at radiation levels of 10⁶ rads/hr, and accumulated dose beyond 10⁶ rads in silicon (⁶⁰Co source). Large format imagers featuring random pixel and subarray addressability, were created for spectroscopy and other scientific applications. They possess a 27 X 27 micrometers ² pixel in 1024_H X 1024_V, 1024_H X 256_V, and 512_H X 512_V formats. Pixels and subarrays (even overlapping subarrays) can be read out destructively or non-destructively. The above features can be combined with 2D on- CID pixel binning because CID binning preserves the spatial fidelity of the pixel charge. Two 1024 linear-type imagers were fabricated with a preamplifier-per-pixel structure and a 27 X 150 micrometers ² large capacity photo-site. One device features on-chip large signal differencing capability between successive exposures. Two 512_H X 512_V (20 X 20 micrometers ² pixel) format imagers were created for UV photon-counting applications. The imagers provide high local count rates through video-rate random subarray addressability and subarray charge injection.

A new family of binary format CMOS CID imagers was designed to meet the random pixel addressing and on-chip signal manipulation requirements of may scientific applications. Key features include true random pixel and programmable subarray addressing, non-destructive readout and charge injection (clearing) that eliminate the need to read out superfluous pixels. And, programmable horizontal/vertical binning provides improved signal/noise and permits spatial signal consolidation even when reading out overlapping subarrays. The imagers incorporate on-chip preamplifiers for low noise readout. Inherent CID pixel characteristics such as non-destructive, non-blooming read-out that permit adaptive exposure control and linear dynamic range extension are maintained. Formats include 1024², 512², and 1024 X 256. All incorporate 27.0 micron contiguous square pixels with in excess of 10⁶ electron well capacity. Serial horizontal and vertical input ports are provided to accept the coordinates of the pixel or subarray to be readout. Rapid subarray readout is facilitated via a single pixel advance clock that is used in conjunction with each random access decoder. A description of the architecture, imager operation and application will be presented.

The INT Prime Focus Mosaic Camera (INT PFC) is designed to provide a large field survey and supernovae search capability for the prime focus of the 2.5 m Isaac Newton Telescope (INT). It is a joint collaboration between the Royal Greenwich Observatory (UK), Kapteyn Sterrenwacht Werkgroep (Netherlands), and the Lawrence Berkeley Laboratories (USA). The INT PFC consists of a 4 chip mosaic utilizing thinned and anti-reflection coated CCDs. These are LORAL devices of the LICK3 design. They will be operated cryogenically in a purpose built camera assembly. A fifth CCD, of the same type, is co-mounted with the science array in the cryostat to provide autoguider functions. This cryostat then mounts to the main camera assembly at the prime focus. This assembly will include standard filters and a novel shutter wheel which has been specifically designed for this application. The camera will have an unvignetted field of 40 arcminutes and a focal ratio of f/3.3. This results in a very tight mechanical specification for co-planarity and flatness of the array of CCDs and also quite stringent flexure tolerance of the camera assembly. A method of characterizing the co- planarity and flatness of the array will be described. The overall system architecture will also be described. One of the main requirements is to read the whole array out within 100s, with less than 10e rms. noise and very low CCD cross talk.

Correlated Double Sampling (CDS) CCD-video signal processing circuits and their application dependent timing. One application is for hollow needle biopsy, a second application is for an instrumentation camera. The last application discussed is for a time delay integrated six CCD sensor large field camera, suing digital CDS.

Two 8 bit successive approximation analog-to-digital converter (ADC) designs and a 12 bit current mode incremental sigma delta ((Sigma) -(Delta) ) ADC have been designed, fabricated, and tested. The successive approximation test chip designs are compatible with active pixel sensor (APS) column parallel architectures with a 20.4 micrometers pitch in a 1.2 micrometers n-well CMOS process and a 40 micrometers pitch in a 2 micrometers n-well CMOS process. The successive approximation designs consume as little as 49 (mu) W at a 500 KHz conversion rate meeting the low power requirements inherent in column parallel architectures. The current mode incremental (Sigma) -(Delta) ADC test chip is designed to be multiplexed among 8 columns in a semi-column parallel current mode APS architecture. The higher accuracy ADC consumes 800 (mu) W at a 5 KHz conversion rate.

CMOS active pixel sensors (APS) allow the flexibility of placing signal processing circuitry on the imaging focal plane. The multiresolution CMOS APS can perform block averaging on-chip to eliminate the off-chip software intensive image processing. This 128 X 128 APS array can be read out at any user-defined resolution by configuring a set of digital shift registers. The full resolution frame rate is 30 Hz with higher rates for all other image resolutions.