Until recently, a solid-state image sensor with full television resolution was a dream. However, the dream of a solid state image sensor has been a driving force in the development of silicon technology for more than twenty-five years. There are probably many in the main stream of semiconductor technology who would argue with this; however, the solid state image sensor was conceived years before the invention of the semi conductor RAM or the microprocessor (i.e., even before the invention of the integrated circuit). No other potential application envisioned at that time required such complexity. How could anyone have ever hoped in 1960 to make a semi conductor chip containing half-a-million picture elements, capable of resolving eight to twelve bits of infornation, and each capable of readout rates in the tens of mega-pixels per second? As early as 1960 arrays of p-n junctions were being investigated as the optical targets in vidicon tubes, replacing the photoconductive targets. It took silicon technology several years to catch up with these dreamers.

The charge-coupled device (CCD) has shown unprecendented performance as a photon detector in the areas of spectral response, charge transfer and readout noise. Recent experience indicates, however, that the full potential for the CCD's charge collection efficiency (CCE) lies well beyond that which is realized in currently available devices. In this paper, we present a definition of CCE performance and introduce a standard test tool (the photon transfer technique) for measuring and optimizing this important COD parameter. We compare CCE characteristics for different types of CODs, discuss the primary limitations in achieving high CCE performance, and outline the prospects for future improvement.

The paper describes the design and performance of a solid-state area-array CCD sensor coupled to a fiber-optic slab. The sensor uses a fra,e transfer organization and has 512 lines of 512 photoelements, each of 18.5 μm x 23.5 μm dimensions, giving, a total active area of 9.472 mm x 12.032 mm. This area being entirely photo-sensitive (no memory zone), the device operates in a single-field mode. The on-chip amplifier allows a readout frequency (pixel frequency) of over 10 MHz. The sensor is coupled to a fiber-optic slab, of unity magnification, overlying the entire photosensitive surface. The elementary fiber diameter is 5.5 μm. A new method of coupling through a gluing process ensures an excellent dimensional stability and immunity to mechanical stresses (vibrations, accelerations ...). The assembly forms a single unit with a metallic centring disk, to facilitate mounting and positioning in equipment. The good performance of the combined sensor and fiber-optic faceplate : low noise (300 μV, wide dynamic range (6500:1) and high illumination response (12 V/μJ/cm2 at λ 550 nm) opens many applications, for example - low-light-level imaging by coupling with an LII tube : nighttime surveillance, astronomy ...) - oscilloscope trace sensing, transient and high-speed phenomena analysis - laser imaging - medical imaging. Furthermore, the 512 x 512 pixel image format is particularly well suited to digital image acquisition and processing. This product was developed as part of a research project for the French Coumissariat A l'Energie Atomique.

In Frame Transfer (FT) imagers all of the imaging area is sensitive to light. If one succeeds in defining pixels of only two photolithographic details (the pixel itself and the separation to its neighbours) FT sensors would allow maximum resolution as well as maximum sensitivity. This paper shows how the use of multiple horizontal read-out registers, vertical anti-blooming and the "accordion principle" allows us to achieve this goal. These concepts are illustrated with a 7.5 mm image diagonal colour image sensor and a 5.2 mm diagonal black and white imager, both with two interlaced fields of 180.000 pixels each.

A review of the present level of CCD technology is presented. Areas in which solid state imaging arrays have become dominant are described. Future areas of application will also be covered.

The features of two-dimensional CCD imaging arrays have led to widespread use of these devices as quantitative optical and x-ray image sensors. In this paper we discuss the desirable attributes that a general purpose digital CCD camera system should have to exploit these capabilities and provide applications flexibility. These include wide dynamic range, flexible readout format and high speed clocking and data acquisition. The ability to digitally sum a number of successive image frames increases the dynamic range and reduces array cooling requirements in intensified-CCD and x-ray applications. We discuss the ways in which these goals have been achieved through the architecture of the DCS-2 Digital Camera System. The system consists of a camera head and controller connected by a datalink over which control information and image data are transmitted. Camera operation is controlled by a Micro-programmable Control Unit (MCU) in response to commands received from the controller over the serial channel in the datalink. The MCU controls all aspects of the CCD readout including integration time and generation of clock signals. The CCD video signal undergoes correlated double sampling and is digitized and transmitted to the controller over the parallel channel in the datalink on a pixel-sequential basis. The instrument controller is based around a microcomputer system with floppy and Winchester disk storage. Successive image frames received from the camera head are summed directly into the Summation Video Memory (SVM). The SVM is a novel three-port design which provides ports for the camera and computer and which also provides an output signal to drive a real-time display monitor. A menu-driven software package provides an interactive environment for control of the CCD readout configuration and image acquisition.

Until recently, the usefulness of the charge coupled device (CCD) as an imaging sensor was thought to be restricted to within rather narrow boundaries of the visible and near IR spectrum. However, since the discovery of backside charging the full potential of CCD performance is now realized. Indeed, the technique of backside charging not only allows the CCD to be used directly in the UV, EUV, and soft X-ray regimes, it has opened up new opportunities in optimizing charge collection processes as well. In this paper, we discuss in considerable detail the technique of backside charging, describing its properties, use, and potential in the future as it applies to the CCD.

This paper describes the performance of two CCD image sensors of 512x512 and 2048x2048 pixel format. These devices were designed specifically for scientific imaging applications.

Fiber-optic coupling of a G.E. TN 2500 CID camera to a microchannel plate detector for vacuum ultraviolet spectroscopy, mass spectroscopy, and various diffraction measurements is discussed. Improvements in resolution of fiber optics and phosphor screens are discussed which result in an increase of greater than 50% in utilization of available CID pixels.

The demanding requirements of Broadcast television have provided a strong driving force for the development of high performance charge-coupled device (CCD) imagers, and, in consequence, there exists today a powerful and general technology base for their manufacture. The demanding requirements of the scientific applications for the charge-coupled-device (CCD) imager, by serendipity, are very well met by this same technology base for the high performance charge-coupled-device (CCD) imager. For example, both TV and astronomy require low noise operation, high transfer efficiency and high resolution, and high quantum efficiency including the blue. As a result the present day manufacturing line dedicated to volume production of high performance charge-coupled-device (CCD) imagers for TV as standard product is also ideally suited to production of high performance scientific charge-coupled-device (CCD) imagers. As discussed in this paper, some key elements of the existing process can now be tailored as required to produce charge-coupled-device (CCD) imagers with some special desired performance characteristics including extended UV response, x-ray or high energy electron response, or enhanced IR response.

We have tested samples of two GEC charge coupled devices with enhanced x-ray performance. The P8607, fabricated with reduced linear dimensions, gives the lowest detector noise reported for Fe-55 with a CCD. The deep depletion device P8600-HR gives the best combination of energy resolution and quantum efficiency at 6 keV. We discuss the energy resolution vs quantum efficiency trade-off curve.

Based on laser holographic technique and volumetric technique a method of generation of hologram, its transmission and display has been described. With reduced bandwidth requirements an approach has been made to extend this technique for implementing a 3-D TV. Using high power laser pulses, this technique involves in registering the holographic images in five CCD planar matrices in a phased manner and in generating video signal from them.

We have developed a photon-counting image acquisition system (PIAS), which can detect a two-dimensional object in a very low-light-level condition. The improvement of the system performance has been accomplished in an image distortion and gain drift owing to the improvement of the silicon PSD (position sensitive detector). The geometric shape of the electrode on the PSD has been redesigned, resulting in the small image distortion of less than 3%. The gain degradation of the electron-bombarded silicon PSD had been a biggest problem in the stable operation of the system. It has been successfully overcome by irradiating the electrons from back side of the PSD instead of front side as formerly. The spatial resolution of the system has been measured to be better than 16 1p/mm typically at the center. Its maximum data is 18 1p/mm or better. Experiments and a theoretical analysis has also been carried out to clarify the limiting parts of the spatial resolution. We have exprimented on system applications in astronomy, spectroscopes and microscopes. In astronomical observation, several kind of stars and galaxies were observed. And also using PIAS combined with the spectroscopy, we measured three spectrums nearby 365nm of Hg-lamp. Further in microscope application, we observed the fibroblast cell of rat by fluorescent microscope using PIAS.

As advanced infrared systems are designed and developed, the requirement for greater sensitivity and resolution is continually encountered. System performance parameters relating to these requirements always depend on the signal-to-noise ratio of the total focal plane. Under optimum operating conditions for the best materials, the response is directly proportional to the total area of sensitive material on the focalplane Ap, whereas the noise varies as Ap1/2 thus system sensitivity varies as Ap1/2. However, resolution depends inversely on the dimensions of the sensitive elements. Thus, higher performance systems will require multielement focal planes of small, closely packed detectors.

Performance characteristics of indium antimonide (InSb) linear infrared (1-5 micrometer) hybrid focal plane assemblies operating in a capacitive-discharge mode (CDM) which use a FET-switched multiplexer are discussed. Specific improvements in each of the two major components of these assemblies have demonstrated increased feasibility for use in high background environments. It is the objective of this paper to detail these improvements and discuss certain additional performance parameters. Previous papers dealt specifically with low background applications using this technique. Emphasis is therefore placed on the characteristics considered significant to the high background, tactical environment. Recent testing of hybrid assemblies using a newly modified Reticon FET-switched multiplexer chip indicate several distinct improvements over the previous version. These include a signal transfer function increase of 25%, a pixel readout rate exceeding 1 megahertz, and a reduction of the kTC component of read noise. The latter resulted in a 13% decrease in the total pixel read noise. These results were accomplished by a general reduction of video line capacitance. Incorporating the detector reset function on the multiplexer chip, thus eliminating the parasitic capacitance of the external MOSFET switch, was the primary factor facilitating the improvement. In addition, a metal bias plate structure has been added to the detector array. This functions to control surface leakage current at large reverse biases. It simultaneously reduces the dark current density and its associated shot noise. This has made it possible to measure maximum integration times of 430 milliseconds at 77 K, an increase of 38% over the previous standard. Large reverse biases are Bow practical and allow extension of the usable charge storage capacity to more than 1.2x108 electrons per storage well.

It is projected that advanced second generation IR systems will use hybrid focal plane arrays consisting of PV HgCdTe detector arrays and silicon CCD signal processor chips. This choice is in concert with the aim of achieving lower NEAT and finer spatial resolution. PV HgCdTe detector arrays have been selected because of power consumption constraints and CCD processor characteristics. In this paper we report on a CCD signal processor chip intended for coupling with PV detector arrays to form advanced second generation IR focal plane arrays. Measurements reveal excellent (4mV RMS) threshold uniformity and the charge division measured noise was less than 200e-.

For staring sensors, improved performance in the location of point targets can be achieved by using an array of hexagonal detectors instead of the usual array of square or rectangular detectors. This fact is demonstrated by calculating the accuracy of the centroid algorithm as a function of signal to noise ratio and blur spot size for both types of detector arrays. The probability density function for the centroid random variable is derived and is used to perform all noise analysis. The analysis indicates that the algorithm error is reduced by as much as a factor of three, the sensitivity to noise is reduced by 17 percent, the computational load is decreased by 23 percent, and the data storage requirement is reduced by 22 percent. The clutter induced noise, as measured by the clutter equivalent target, is essentially identical for square and hexagonal detectors of the same area.

In this paper D* is introduced, and analyzed as a figure of merit for focal plane arrays. Some of the inadequacies of D* are presented, and used to motivate the use of the signal to noise ratio as a figure of merit. Spatial noise (nonuniformity) is incorporated into the signal to noise ratio and its effects are examained for the cases of a Schottky barrier, and standard photon detector (constant quantum efficiency) over the 3 to 5 μm spectral band.

We demonstrate that it is straightforward to fabricate a mosaic array of hexagonal detectors within the framework of existing design rules and fabrication technology. As an example, we have fabricated an 8x8 array of hexagonal detectors with a very high fill factor using a PtSi technology. This is believed to be the first time that such an array has actually been fabricated. Design and fabrication highlights are discussed. A brief summary of performance advantages and algorithm implications that result from the use of the hexagonal configuration is given along with a possible system configuration. Plans for the design and testing of a 64x64 array of hexagonal detectors and corresponding readout chip are presented.

Performance measurements of two Multispectral Linear Array focal planes are presented. Both pushbroom sensors have been developed for application in remote sensing instruments. A buttable, four-spectral-band, linear-format charge coupled device (CCD) and a but-table, two-spectral-band, linear-format, shortwave infrared charge coupled device (IRCCD) have been developed under NASA funding. These silicon integrated circuits may be butted end to end to provide very-high-resolution multispectral focal planes. The visible CCD is organized as four sensor lines of 1024 pixels each. Each line views the scene in a different spectral window defined by integral optical bandpass filters. A prototype focal plane with five devices, providing 4x5120-pixel resolution has been demonstrated. The high quantum efficiency of the backside-illuminated CCD technology provides excellent signal-to-noise performance and unusually high MTF across the entire visible and near-IR spectrum. The shortwave infrared (SWIR) sensor is organized as two line sensors of 512 detectors each. The SWIR (1-2.5 μm) spectral windows may be defined by bandpass filters placed in close proximity to the devices. The dual-band sensor consists of Schottky barrier detectors read out by CCD multiplexers. This monolithic sensor operates at 125°K with radiometric performance. A prototype five-device focal plane providing 2x2560 detectors has been demonstrated. The devices provide very high uniformity, and excellent MTF across the SWIR band.

Thermal imaging in the 3-5 micrometer spectral band using Pt-Si Schottky barrier mosaic focal planes has been reported extensively. The content of these papers leaned heavily toward military applications of cameras developed using these devices. The thermal imaging cameras based on the use of these arrays are simple to operate and provide excellent imagery without the need for complex and expensive image correcting or processing. Thus, staring cameras utilizing Schottky photodiodes are low cost. The state-of-the-art in silicon technology provides ruggedness and reliability in these devices making them readily affordable for a variety of non-military applications. We will touch on some of the potential uses for this type of camera and show photographs of typical imagery obtained both day and night, indoor and outdoor.

A novel, fully programmable, parallel-architecture, analog array processor will be described which can meet the high throughput requirements of real-time image processing systems. This architecture, when implemented with charge coupled device technology, can be interfaced directly with a PtSi sensor array to provide the capability of fabricating a compact, high-throughput module which has application in a wide variety of areas in tracking, guidance and position sensing.

We present a description of the design and performance of an imaging spectrometer intended for operation in the seven to fourteen micrometer wavelength range. It is based upon a sixteen by sixteen element Si:Bi hybrid array, a circular variable filter wheel, and a microcomputer data acquisition system. The spatial-spectral photometric performances have been evaluated in the laboratory as well as field tested at astronomical observatories. The performance characteristics of the system are presented as well as laboratory and astronomical images.

In this paper we describe the construction of an infrared imaging camera system utilizing a linear 32-element InSb array developed by Cincinnati Electronics (CE) and the Jet Propulsion Laboratory (JPL). This camera system was designed for astronomical infrared imaging in the J, H, K, L, and M bands, primarily at the Cerro Tololo Interamerican Observatory (CTIO)* in Chile. Previously published reports 1,2,3,4,5,6 indicated that this array had the desirable characteristics required for astronomical applications, such as high quantum efficiency, low dark current with lower temperature operation, and large well capacity. The only drawback was the relatively high noise resulting from the high video line capacitance. In fact, the noise has proven to be the most difficult system problem because low-background operation necessitates integration times sufficient to insure that source flux noise becomes the limiting factor.