In the early and mid-1970s, digital image processing was the subject of intense university and corporate research. The research lay along two lines: (1) developing mathematical techniques for improving the appearance of or analyzing the contents of images represented in digital form, and (2) creating cost-effective hardware to carry out these techniques. The research has been very effective, as evidenced by the continued decline of image processing as a research topic, and the rapid increase of commercial companies to market digital image processing software and hardware.

As a basis for calculating lens correction curves, the front and rear focal and nodal surfaces are mapped for lenses used in film analysis. The image of a collimated source is read as a function of incident angle using a linear CCD. The front nodal point is used as a camera position and the measured angular variation provides a fine correction td the object distance. The resulting corrections allow accurate spatial locations to be calculated for target objects imaged at the film plane.

A new model of sperm swimming kinematics, which uses signal processing methods and multivariate statistical techniques to identify individual cell-motion parameters and unique cell populations, is presented. Swimming paths of individual cells are obtained using real-time, video-edge digitization. Raw paths are adaptively filtered to identify average paths, and measurements of space-time oscillations about average paths are made. Time-dependent frequency information is extracted from spatial variations about average paths using harmonic analysis. Raw-path and average-path measures such as curvature, curve length, and straight-line length, and measures of oscillations about average paths such as time-dependent amplitude and frequency variations, are used in a multivariate, cluster analysis to identify unique cell populations. The entire process, including digitization of sperm video images, is computer-automated. Preliminary results indicate that this method of tracking, digitization, and kinematic analysis accurately identifies unique cell subpopulations, including: the relative numbers of cells in each subpopulation, how subpopulations differ, and the extent and significance of such differences. With appropriate work, this approach may be useful for clinical discrimination between normal and abnormal semen specimens.

In the field of cell motility, computer assisted studies have become a necessity both for scientists studying bacteria and scientists studying more complex cells (1). However, the needs of the scientist differ depending upon the organism being studied. For instance, most bacteria and ciliated or flagellated protozoans do not change shape when moving or when making decisions on directionality in chemical gradients. Therefore, in these cases, systems which track the centroid of the cell are sufficient for most studies. In marked contrast, most animal cells ranging in complexity from slime mold amebae to human white blood cells are ameboid and change shape both when locomoting and when making decisions about directionality. For these latter cells, simply tracking the translocation of the centroid can be misleading, as will become evident in the discussion which follows. To assist in the analysis of ameboid cell motility, we developed a software package for the motion analysis machines (The Expert Vision System and the Cell Track System) marketed by Motion Analysis Corporation of Santa Rosa, California. This system is referred to as the "Dynamic Morphology System" and is tailored for use by cell biologists interested in two dimensional studies of the temporal dynamics of cell shape, cell extensions (pseudopods) and cell contractions (1,2,3,4,). It is also unusually well suited for assessing the effects of mutation (1) on cell behavior and chemotaxis (the directed movement of cells in chemical gradients) and for comparing normal and cancer cells. In the discussion which follows, I will describe some of the capabilities of the system and its application to the problem of cell movement and ameboid behavior in chemical gradients. I will also briefly describe the development of a 3D Dynamic Morphology System and a system for acoustically analysing complex data generated by the Dynamic Morphology System.

Three-dimensional human motion analysis has been used for complex kinematic description of abnormal gait in children with neuromuscular disease. Multiple skin markers estimate skeletal segment position, and a sorting and smoothing routine provides marker trajectories. The position and orientation of the moving skeleton in space are derived mathematically from the marker positions, and joint motions are calculated from the Eulerian transformation matrix between linked proximal and distal skeletal segments. Reproduceability has been excellent, and the technique has proven to be a useful adjunct to surgical planning.

We describe the use of high speed video to provide quantitative assessment of motion in athletic performance. Besides the normal requirement for accuracy, an essential feature is that the information be provided rapidly enough so that it my serve as valuable feedback in the learning process. The general considerations which must be addressed in the development of such a computer based system are discussed. These ideas are illustrated specifically through the description of a prototype system which has been designed for the javelin throw.

Special video path editing capabilities with custom hardware and software, have been developed for use in conjunction with existing video acquisition hardware and firmware. This system has simplified the task of quantifying the kinematics of human movement. A set of retro-reflective markers are secured to a subject performing a given task (i.e. walking, throwing, swinging a golf club, etc.). Multiple cameras, a video processor, and a computer work station collect video data while the task is performed. Software has been developed to edit video files, create centroid data, and identify marker paths. Multi-camera path files are combined to form a 3D path file using the DLT method of cinematography. A separate program converts the 3D path file into kinematic data by creating a set of local coordinate axes and performing a series of coordinate transformations from one local system to the next. The kinematic data is then displayed for appropriate review and/or comparison.

A complete system to quantify three-dimensional human movement kinematics was developed and initial accuracy testing was completed. This video-based system incorporates a video processor produced by Motion Analysis Corporation to edge-detect markers in a calibrated space, a SUN computer work station to receive data from the video processor, and appropriate software for analysis of the data. A series of tests was performed to compare static and dynamic accuracy of this video-based system to that reported of other video and film-based systems. The mean absolute errors for the static testing were 4.0 mm in the X direction, 2.5 mm in the Y direction, and 5.4 mm in the Z direction, with standard deviations of 4.0, 1.8, and 5.9 mm respectively. The mean absolute errors for the static testing were 11.1, 2.6, and 17.6 mm in the X, Y, and Z directions, respectively, with standard deviations of 0.4, 0.5, and 1.1 mm. These errors are comparable to those of other video-based systems but are slightly higher than those found with film-based systems.

For long-pulse (150 microseconds), side-pumped lasers, such as dye or reactor-pumped lasers, a significant transverse index gradient can develop. In order to design an appropriate laser resonator for efficient power extraction, a detailed space- and time-resolved determination of the index perturbation is required. We have applied high-speed photography to the measurement of wavefront errors in a pulsed laser gain region using a fast framing camera. We utilized both a derivative of a Hartmann test technique and a shearing interferometer, with good agreement between the two. The measured wavefronts have a predominantly parabolic index shape except near the walls, with a roughly linear time dependence during the pumping pulse.

The versatility of the LLNL developed 75-mm format high-speed electro-optic camera has been improved by the addition of a microchannel plate image intensifier to the electronic image converter, and by the replacement of the discreet objective lens with a commercially available zoom lens. The image intensifier can be operated at a gain of up to 10,000. This allows the use of nanosecond or subnanosecond laser strobe light pulses for object illumination thus improving the time resolution of the camera by at least an order of magnitude. The discreet objective lens of the camera has been replaced with a fast (f/4) 200-400 mm zoom lens that when coupled with a quick-connect telephoto adapter (either 1.4x or 2.0x) allows the camera to operate at any focal length from 200 to 800 mm. This permits variable magnification of the object without having to move the camera and without any degradation of dynamic spatial resolution or significant loss of light gathering ability.

A solid state light control system was designed and fabricated for use with hydrodiagnostic streak cameras of the electro-optic type. With its use, the film containing the streak images will have on it two time scales simultaneously exposed with the signal. This allows timing and cross timing. The latter is achieved with exposure modulation marking onto the time tick marks. The purpose of using two time scales will be discussed. The design is based on a microcomputer, resulting in a compact and easy to use instrument. The light source is a small red light emitting diode. Time marking can be programmed in steps of 0.1 microseconds, with a range of 255 steps. The time accuracy is based on a precision 100 MHz quarts crystal, giving a divided down 10 MHz system frequency. The light is guided by two small 100 micron diameter optical fibers, which facilitates light coupling onto the input slit of an electro-optic streak camera. Three distinct groups of exposure modulation of the time tick marks can be independently set anywhere onto the streak duration. This system has been successfully used in Fabry-Perot laser velocimeters for over four years in our Laboratory. The microcomputer control section is also being use in providing optical fids to mechanical rotor cameras.

An analysis is presented that specifies the corrections necessary when velocity interferometic measurements are made of surfaces viewed through shocked window materials. The analysis is performed in a Lagrangian (material) coordinate system, is limited to one-dimensional planar, cylindrical and spherical flow conditions, assumes the index of refraction, n, depends on the density, p, alone and is taken in the classical limit. In all cases of cylindrical and spherical geometry and for unfavorable n(p) dependencies in planar geometry, the relationship between the apparent and true interface velocities includes a term that must be integrated through the flow from the interface to the shock front. Determination of the true velocity history of an arbitrary wave thereby requires an iterative procedure of working the entire hydrodynamic problem. It is shown that the integral terms are eliminated by planar geometry and a refractive index dependence με = μ0 -k(1-ε) where μ = n-1, ε = po/p, the sub-o denotes the unshocked state, and k, is a constant. Results from Fabry-Perot velocity interferometer measurements are presented to demonstrate various predictions of the analysis and to indicate that several alkali halides exhibit the special refractive index property cited above.

A streak camera was used to image moving interference fringes from the probe volume of a proposed laser induced fluorescence velocimeter, permitting easy visualization of the light intensity as a function of space and time and providing a method for directly measuring the fringe velocity.

The need to study the effects of the impact of micrometeorites and orbital debris on various space based systems, has brought together the technologies of several companies and individuals in order to provide a successful instrumentation package. A light gas gun was employed to accelerate small projectiles to speeds in excess of 7 kilometers per second. Their impact on various targets is being studied with the help of a specially designed, continuous access, rotating mirror framing camera. The camera provides 80 frames of data at up to one million frames per second with exposure times of 20 nanoseconds.

The Combination Camera described here is a traditional rotating-mirror streak camera with several modern improve-ments, including an optional image intensifier camera back for use with weak light sources. It is intended for use in routine testing of high explosive devices, and for experimental studies of detonation phenomena. The mechanical design is carefully thought out to make precision alignment of the camera to the subject easy, and to make focusing quick and accurate. Interchangeable slit mountings provide single, multiple, and special slits that are accurately aligned to maximize the useful information from an experiment. The image intensifier provides light gains up to 5,000, so that sources with equivalent blackbody temperatures as low as 1500 K can be recorded with exposure times as small as 5 ns. The length of a single slit is 28 mm, and the total height of a multiple slit pattern can be as much as 28 mm. The film back for use without intensification uses 70-mm roll film, and the film back for use with the intensifier uses 4x5 film. Measured resolution on film is 60 1p/mm without intensification, and 16 1p/mm with intensification at maximum gain, and 18 1p/mm at lower gain. The camera has been found especially useful for laser interferometric velocimetry using a Fabry-Perot interferometer as the dispersing element, where light gain is usually necessary even with the largest lasers as light sources.

A new microfocus X-ray source has been developed specifically designed for high speed cine radiographic applications. The combination of small X-ray source size, high output flux and the ability to electronically gate the X-ray output for short time periods makes the system particularly suitable for use with an X-ray image intensifier and image converter camera combination. Using this system dynamic events have been recorded at framing rates up to one million frames per second.

This report provides information to professional industrial, scientific, and technical photographers as well as research personnel on filtration with glass and infrared-absorbing and -reflecting filters. Glass and infrared filtration is a simple and effective method to reduce the radiation heat transfer associated with continuous high-intensity tungsten lamps. The results of a filtration experiment conducted at the NASA Lewis Research Center in Cleveland, Ohio, are explained. The figures provide starting points for quantifying the effectiveness of various filters and associated light intensities. The combination of a spectrally selective reflector (hot or cold mirror) based on multilayer thin-film prin-ciples and heat-absorbing or infrared opaque glass results in the maximum reduction in temperature rise with a minimum of incident light loss. The report recommends use of a voltage regulator to further control temperature rise and incident light values.

This paper deals with two applications of streak photography in everyday industrial and biological research and development endeavors. In each instance, the simplicity and low cost with which the data could be reduced to useful information and the ease with which it could be communicated to other technical personnel was an important side benefit. In the first case, a laboratory built rotating drum streak camera with a 12" circumference, was used to study the performance of the new Olympus F280 flash on an Olympus 0M4-T camera. The strobing frequency, starting characteristics, duration under manual and camera control, flash relationship to shutter curtain position, flash initiation and curtain velocities were examined. In the second case, a standard oscilloscope recording camera was modified to study the growth characteristics of a fungus culture growing under controlled laboratory conditions for periods up to one full week. The streak photographs, taken at ultra slow rates, were correlated with 2-dimensional photographs taken at regular intervals to eliminate ambiguity in the growth cycle whe the photographs were included in a written report.

The mathematical solution for the true positions of features in space from the position of their images on pairs of radiographs has been discussed in Ballistic Research Laboratory reports by Grabarek and Herr (BRL TN 1634, September 1966, AD 807619), and by Henry C. Dubin (BRL MR 2470, April 1975, AD B003797L, now available for unlimited distribution). A truly general solution should not require, or be limited to, orthogonality of the image pairs. Dubin provides such a solution. Because of errors in setup and measurement, the two lines presumed to connect the respective source and image points through the common object point do not necessarily intersect. Using vector notation and partial derivatives, he obtains the line of minimum length between these two vectors. The midpoint of this line is the best estimate for the object position, and the length is a measure of the error of the estimate. The redundant image coordinate in Dubin's method contributes to increasing the accuracy of the estimate of the position of features. The solution of Grabarek and Herr uses analytic geometry and the assumption that the two lines between the respective sources and images intersect, and requires an orthogonal radiographic setup. This approach forgoes generality and some available accuracy. Driven by the need to provide as simple an approach as possible, this paper presents two similar derivations. The author uses analytic geometry to rederive the equations of Grabarek and Herr from a simpler perspective. The form of the solution provides a conceptual bridge to a more direct derivation by the author, using trigonometry. Both derivations are for orthogonal radiographic setups. The trigonometric approach does not require complicated computation of magnification factors, is more easily understood in terms of the geometry of the setup, and is easily implemented in computer or calculator programs to reduce orthogonal radiographs.

Although modern numerical computational techniques are able to predict the geometry of a complex detonation front in condensed explosives, in most instances, experimental measurements are still required to verify computational results. We have measured non-planar detonation fronts by means of flash radiography as well as multi-slit ultra-high speed streak photography. In this paper, we shall discuss the two techniques and show measurement results from the two techniques.

The use of flash x-ray radiography to investigate high speed multiphase flows is demonstrated. Both fundamental and practical problems of interest to the pulp and paper industry are used as examples. More specifically, studies of concentrated fiber suspension flows, kraft black liquor sprays and impulse drying are discussed.

Flash radiography has been used for many years to measure ballistic phenomena and explosive events which are impossible to capture with high speed photography due to the large amount of smoke and flash usually associated with these events. The basic system consists of a pulse generator, high voltage power supply, and a cold cathode field emission x-ray tube. This paper will deal with the exposure fields around 150, 180, and 300 kV Hewlett Packard x-ray tubes. Although the dosage of each of these systems is described in the system characteristics, it is described only in terms of an exposure at one given distance for each energy level, i.e., 40 mR at 20cm from the 150 kV tubehead. This paper will examine from a safety standpoint, the exposure obtained anywhere within a 360° field and will compare these results with the yearly dosage permitted by the U.S. government. This study will supply quantitative data to aid in determining the adequacy of standard operating procedures for flash x-ray equipment.

The flash radiography combining a low impedance high intensity flash x-ray generator and a digital radiography system for biomedical use is described. This flash x-ray generator was a single shot type and consisted of the following essential components: a high voltage generator, a low impedance transmission line using an oil condenser of 120kV-0.15pF, a high speed impulse switching system with a time resolution of less than 1ps, and a new type of x-ray tube having two sets of electrodes. This source could be used for condenser charging voltages of 50 to 120kV and peak currents of 20 to 40kA. The maximum intensities for the Type A and B electrode combinations were about 20 and 40C/kg at lm/pulse, respectively. The exposure times were about fps . The optimum output corresponding to the optimum x-ray quality could be controlled. The peak intensity of the spectrum distribution varied from a low to high photon energy region, and the energy latitude of this spectrum also varied from narrow to wide. The size of the effective focal spot ranged from 0.2 to 3.0mm in diameter. Various kinds of biomedical flash radiography (e. g., the continuous delayed imaging, and image analysis equivalent to the three dimensional image analysis) were accomplished by controlling the source conditions concerning the x-ray intensity, the quality, the spectrum distribution, the effective focal spot, and the delay time.

In order to accurately predict the performance of a kinetic energy (KE) penetrator against an oblique target plate, a clear understanding of the interaction between the penetrator and target is required. Critical to the analysis and the formulation of any theory to predict this interaction is the accurate measurement of the rate of change of the penetrator's length, mass, and yaw angle during penetration and immediately following its exit from the plate. An experimental technique was adopted in which flash radiography was used to measure penetrator characteristics as a function of time. This paper presents the conventional experimental technique, the improved experimental technique, and four test arrangement considerations associated with the multiple image radiography of a single penetrator. A limited number of quarter scale experiments of tungsten rod penetrators versus one inch aluminum (7075-T6) plates at 60° obliquity were performed and analyzed using the improved technique. Flash radiography was used to capture multiple images of the residual penetrator over a 100 OKs time window after it fully perforated the target plate. This data will be used to develop new predictive equations for rod penetration.

High speed flash radiography has been used during the past few decades to record phenomena that occur during rapid dynamic events. The events are difficult, if not impossible, to record by other means due to the speed of the event or the obscuration associated with it. To eliminate the motion blur of objects moving in excess of 6,000 meters/sec (20,000 ft/sec), it is necessary to have an extremely short exposure time (5 to 70 nanoseconds). This short exposure time requires the use of high speed intensifying screens and high speed x-ray film to record the radiographic image. Technicians who use flash x-rays have to depend on recommendations from present and former flash x-ray users for film and screen selection. The film and screen industry has made many changes in the last few years. It is not uncommon to find that the particular film or screen you wish to use is no longer manufactured. This paper will describe some of the films and screens that are currently manufactured. It will describe their relative exposures when used with the Hewlett/Packard model 43731A 150 kV Flash X-ray System.

This paper describes a cineradiography system in use at the Los Alamos National Laboratory, as related to the advantages and disadvantages over conventional flash x-ray systems. Traditionally, x-ray imaging techniques in dynamic testing have relied on creating extremely short pulses of radiation to freeze the motion of the object, and then recording the image on film by means of fluorescent intensifying screens to obtain sufficient image density on the film. This results in images often limited only by the resolution of the film-screen combination, which are usually of reasonable quality. In a cineradiography system, two basic differences are evident. First, the radiation source emits continuously for the duration of the experiment. Second, the film is replaced by a gated, intensified television camera focused on the fluorescent screeen. The image is frozen by the short gate time of the camera, rather than by the short pulse of radiation. One advantage of the television system is that the camera can be considerably distant from the screen, and if the screen is sacrificial, mechanical protection requirements are alleviated or eliminated. Another advantage is that several cameras can be focused on the same screen, allowing multiple images to be made with the same geometry. A third advantage is that the spot size of the radiation source is small, thus reducing geometrical limitations on resolution. The disadvantages of this system relate to the use of the television camera to record the image(s). Neither the resolution nor the contrast of the intensified television camera is as good as film, and this limits the quality of the image that can be produced. However, flash radiographs are often of relatively poor quality because of the limited amount of radiation available from the source and the graininess of the high-speed film required, so this is often not an important difference.

The fundamental studies for a repetitional (triple exposure) pulsed x-ray generator having variable spectra for biomedical imaging are described. This generator consisted of the following components: a high voltage generating unit, a voltage divider unit, three high voltage pulsers with output voltages ranging from 50 to 200kV, a triple parallel type of high speed impulse switching system, a high power gas diode having three terminals, and two types of evacuated x-ray tubes, each of a different type. The pulse condensers of the pulsers were charged to the same or different energies by using a voltage divider unit and were connected to the single x-ray tube through a high power gas diode. The duration of each x-ray pulse was about 1ps, and the minimum time interval between two pulses was about 100ps. The maximum intensity was about 10C/kg at lm/pulse, and the effective focal spot size varied from 0.2 to 3.0mm in diameter. The triple exposure of pulsed x-rays having variable spectra and time intervals was obtained by controlling the condenser charging voltage, the tube impedance, the diode impedance, and others.

This tube is aimed at shuttering speeds faster than 200 picoseconds, with required electrical gate pulse less than 80 volts. Similar tubes were described in a 1984 paper at this meeting. One new feature is the use of a microchannel plate (MCP) to provide high and adjustable signal gain. This tube, in comparison with proximity-focussed tubes, conserves resolution by using an electrostatic focus lens. Also new is the use of transfer photocathode technique to eliminate stray emission from the gate mesh. Measured performance: Gating speed, Gain, On-Off signal ratio will be presented. Tube construction and the transfer photocathode technique will be discussed. The noise which shows up in high-gain imaging systems will be discussed.

A pulsed laser is used as a light source for illuminating the surface of a dynamic event of <1 mm 2 moving at >3 mm/μs. At a predetermined time during the dynamic action, a stereo camera is used to record a pair of images of the dynamically moving surface. The stereoimage pair can be quantified for surface contour.

Several new materials have electrooptical coefficients that are high in comparison with that of lithium niobate (31 x 10-12 m/V). These new materials are SrxBai,Nb206 (SBN:100x); Sr Ba 1 K1 Na Nb206 (BSKNN); and polymeric, ultrathin films. For SBN:60, SBN:75, and BSKNN, electrooptical coefficients of 400 x 10- m7 V, 1400 X 10 m/V, and 150 x 10-12 m/V, respectively, have been reported.' In this paper, we review the theory of light-steering machines and predict the theoretical performance of these systems using these new electrooptical materials. We include new, innovative methods for increasing the resolution of these systems. In addressing problems in fabricating the crystals, we discuss crystal properties that must be controlled if the development of a practical crystal streak camera is to be successful. Finally, we present our experimental results from a laboratory-built light deflector using SBN:60.

A two-source shear pattern recording is proposed as a method for single-shot measurement of the pulse shape from nearly monochromatic sources whose pulse lengths are shorter than their coherence times. The basis of this method relies on the assertion that if two identical electromagnetic pulses are recombined with a time delay greater than the sum of their pulse widths, the recordable spatial pattern has no fringes in it. At an arbitrary delay, translated into an actual spatial recording position, the recorded modulated intensity will sample the corresponding laser intensity at that delay time, but with a modulation due to the coherence function of the electromagnetic pulse. Two arrangements are proposed for recording the pattern. The principles, the design parameters, and the methodologies of these arrangements will be presented. Resolutions of the configurations and their limitations will be given as well.

The paper covers the basic differences between the traditionally used S.I.T. detector and a currently available CCD detector. Results are shown from a commercial CCD based 2-dimensional 12-bit readout system. Data is shown from a simulation of multipoint recording from a fibre optic array at high time resolution is shown along with an evaluation of the same system applied to a framing camera.

When a transient process of non-reproducible enents is studied by the pulsed holography, the effective approach is that the multi-holograms or multi-holographic interferograms can be recorded sequentially in single transient process. For this purpose we researched the electro-optical framing type multipulsed holography and made a excellent internal controlling electro-optical framing device successfully. An in-line light pulses from ruby laser can be resolved into spatial multiplexing beams with desirable direction that form several reference beams by the framing device. Based on the research of this technique, the High-Speed Holocamera with Internal-Controlling Electro-Optical Framing Device is made. It consists of a multi-pulsed single frequency ruby laser, a set of electro-optical framing device and an electronic controlling system. It can record continuously 4 holograms or 4 interferograms in a dynamic process.

In this paper we describe the operation and applications of a commercially available framing camera capable of four separate two-dimensional images with each frame having a 120-ps gate width. The time separation between frames can be selected arbitrarily. Fast gating of a single frame is accomplished by using a wafer image intensifier tube in which the cathode is capacitively coupled to an external electrode placed outside of the input window of the tube. This electrode is then pulsed relative to the microchannel plate by a narrow (120 ps), high-voltage pulse. The transmission of the tube as a function of time has been verified using a laser-diode pulser and is approximately Gaussian with a FWHM of 120 ps. Multiple frames are obtained by using multiple gated tubes which can then share a single bias supply and pulser with relative gate times selected by the cable lengths between the tubes and the pulser. A beamsplitter system has been constructed which produces a separate image for each tube from a single scene. In the present system, the tubes use S-20 photocathodes with an 18 mm diameter and quartz input windows. Spatial resolution is unchanged between d.c. and fast gated operations and has been measured to 10 1p/mm. Applications to time-dependent behavior will be presented.

A conceptually new sampling optical oscilloscope has been developed based on the streak camera technology. This is analogous to a conventional sampling oscilloscope but entirely electron-optical in operation and can potentially analyze picosecond optical signals directly. The sampling optical oscilloscope makes use of a photoelectron tube so-called sampling streak tube. This tube is not only a photodetector but also an ultrafast electron sampling device. Optical signals are converted into Photoelectrons, which is the replica of the incident optical signals, at the photocathode. These photoelectrons are then accelerated and deflected to perform traversing of the replica for the purpose of sampling. Part of the electron replica is incident on a Phosphor screen, where it is converted into an optical signal. A ohotomultiplier tube is used to detect the radiation of the screen to analyze the incident signal. The entire waveform of the input signal is obtained by gradu-ally changing the deflection or sampling timing. This unique system permits direct detec tion, digitization and analysis of optical signals in the 350-850 nm region with rise and fall time of 10 ps. It provides a dynamic ranee of > 1000:1, sampling rate of up to 2 MHz, sensitivity of photomultipliers, and breaks the distortion such as overshoot, rincino, cable mismatches or other instrument distortion.

The accurate characterization of high-power multi-beam laser systems requires an optical streak camera system having a linear dynamic range of 1000. In this paper we wish to show that the classic definition of streak camera dynamic range tells very little about the system linearity. A problem of flare in the recording system limits the linearity to < 100. Data from two commercial streak camera systems is presented.

We have completed preliminary characterizations of a new 3-grid, large-format, ITT1 streak camera tube operating with about a 2.3-kV extraction voltage. We have found that the best attainable impulse response is 10 ps at low input intensity. However, the impulse response increases to only 30 ps with an increase of nearly three orders of magnitude in input intensity. Further, the spatial resolution remains greater than 5 line pairs/mm at 50% contrast over these three orders of magnitude of input intensity. We have also done preliminary computational modeling of the tube's temporal response as a function of focus voltage, illumina-tion spot size, and wavelength.

This paper describes the methods used by EG&G Energy Measurements to characterize a streak tube. Both radiometric (QE, uniformity, gain, linearity, extinction ratio, opacity) and resolution (static, spatial, and temporal) tests are covered. The signif-icance of these measurements with respect to streak camera applications is discussed. The results of the measurements on a streak tube designed and built by EG&G Amador Valley Operations are included as an illustration.

Using high speed photography with image conveter to diagnose superspeed phenomena is an important way. In this paper the author developed the approximate formulae of the temporal and spatial modulation transfer functions of a streak image tube.

The RCA C73435 image tube is biased at voltages far from its original design in the LLNL ultrafast (10 ps) streak camera. Its output resolution at streak camera operating potentials has been measured as a function of input slit width, incident-light wavelength, and focus-grid voltage. The temporal resolution is insensitive to focus-grid voltage for a narrow (100 pm) input slit but is very sensitive to focus-grid voltage for a wide (2 mm) input slit. At the optimum wide-slit focus voltage, temporal resolution is insensitive to slit width. Spatial resolution is nearly independent of focus-grid voltage for values that give good temporal resolution. Both temporal and spatial resolution depend on the incident-light wavelength. Data for 1.06-μm light show significantly better focusing than for 0.53-μm light. Streak camera operation is simulated with a computer program that calculates photoelectron trajectories. Electron ray tracing describes all of the observed effects of slit width, incident-light wavelength, and focus-grid voltage on output resolution.

The LLNL optical streak camera is used by the Laser Fusion Program in a wide range of applications. Many of these applications require a large recorded dynamic range. Recent work has focused on maximizing the dynamic range of the streak camera recording system. For our streak cameras, image intensifier saturation limits the upper end of the dynamic range. We have developed procedures to set the image intensifier gain such that the system dynamic range is maximized. Specifically, the gain is set such that a single streak tube photoelectron is recorded with an exposure of about five times the recording system noise. This ensures detection of single photoelectrons while not consuming intensifier or recording system dynamic range through excessive intensifier gain. The optimum intensifier gain has been determined for two types of film and for a lens-coupled CCD camera. We have determined that by recording the streak camera image with a CCD camera, the system is shot-noise limited up to the onset of image intensifier nonlinearity. When recording on film, the film determines the noise at high exposure' levels. There is discussion of the effects of slit width and image intensifier saturation on dynamic range.

To accurately measure pulse amplitude, shape, and relative time histories of optical signals with an ontical streak camera, it is necessary to correct each recorded image for spatially-dependent gain nonuniformity and geometric distortion. Gain nonuniformities arise from sensitivity variations in the streak-tube photocathode, phosphor screen, image-intensifier tube, and image recording system. These nonuniformities may be severe, and have been observed to be on the order of 100% for some LLNL ontical streak cameras. Geometric distortion due to optical couplings, electron-optics, and sweep nonlinearity not only affects pulse position and timing measurements, but affects pulse amplitude and shape measurements as well. By using a 1.053-μm, long-pulse, high-power laser to generate a spatially and temporally uniform source as input to the streak camera, the combined effects of flat-field response and geometric distortion can be measured under the normal dynamic operation of cameras with S-1 photocathodes. Additionally, by using the same laser system to generate a train of short pulses that can be spatially modulated at the input of the streak camera, we can effectively create a two-dimensional grid of equally-spaced pulses. This allows a dynamic measurement of the geometric distortion of the streak camera. We will discuss the techniques involved in performing these calibrations, will present some of the measured results for LLNL optical streak cameras, and will discuss software methods to correct for these effects.

There have been many papers to describe the electro-optical systems of the image converters focused with parallel electric and magnetic fields.`"" Some problems appeared when it is used in a PS scan image converter are explored and the feature that the system can not be used in a even pitch focusing is presented in this paper. The results of the dynamical measurements using solid laser for a sample tube are also given.

In the fast pace of the automotive industry the ability to make sound decisions rapidly can be extremely cost effective. Often these decisions involve some important changes to a product in the final development stage or early production stage and are a result of testing the product. The use of high speed video can be a valuable tool in many areas of research and engineering testing because of the ability for instant review. This can greatly reduce the decision making time lag. This paper will describe the use of a high speed video system for automotive barrier safety testing. This application involves using the system to determine the rearward horizontal displacement of the steering column. There is a Federal Motor Vehicle Safety Standard which sets a limit on the rearward motion of the steering column. Having the ability to perform this analysis and get reliable results quickly has proven to be helpful in timely decision making.

Several new photoconductors now commercially available as targets in Type 7803 Focus Projection and Scan (FPS) electrostatically focused vidicons have been characterized for use as radiometric sensors in transient illumination and single frame applications. These include Saticon (Se + As + Te), Newvicon (ZnSe), Pasecon (CdSe), and Plumbicon (PbO). Samples from several domestic and foreign manufacturers have been evaluated for photocon-ductive response time and responsivity at selected narrow wavelength bands, including 410 nm, 560 nm, and 822 nm. These data are compared with performance data from older target materials including Sb2S3 and silicon. The effect of bias lighting on sensitivity and transfer curve slopes for single-event pulsed light stimulas are presented. Dynamic range and saturation limits as functions of beam aperture, target voltage, and filament current are also discussed.

This paper describes a method for controlling light level for an intensified focal plane array low light level television camera. Light integrating properties of the image intensifier photocathode, and the focal plane array (FPA) are exploited to control light exposure by time domain control only, gating the intensifier photocathode. Shutter efficiency effects of charge propagation velocity on the photocathode are empirically examined.

The introduction of a new high speed video system in the Spring of 1987 doubled the rate at which color video images could be recorded. In addition to the 400 picture per second sampling rate, advances in image sensor technology resulted in a unit with high light sensitivity increasing the number of applications were video instrumentation could be applied. A device called a "Wave Inserter' was also introduced, which allowed an electronic signal to be displayed over the video picture in a manner analogous to oscillography with 16mm rotating-prism. cameras. Providing a built in digital memory for picture storage lets users easily detect subject displacement in video images. The purpose of this paper is to review a number of applications where the new hardware results in better image acquisition, and to provide a detailed description of the equipment and its capabilities.

A useful and foolproof method of analyzing explosive phenomena and designing blasts is with information obtained from high-speed, 16mm motion picture photography. Utilizing this information along with the methodologies developed adds confidence to custom and site specific blast designs for use in mining, quarry, coal, utility, construction, research and specialized military applications. The numerical and analytical techniques developed allow a blaster or engineer to evaluate blasting performance, compare explosive efficiencies, optimize millisecond delay timing between detonations, evaluate damage to particular seams of material such as coal, optimize overburden casting, predict explosive energy required to displace toe burdens, compare over-all blast designs, investigate air decks for presplitting and reclamation purposes, and to study general detonation phenomena. Results are based on the analysis of a few hundred films of blasts performed in full-scale production environments. Instrumentation, field set-ups, and specific data interpre-tation are discussed in detail. The presentation is also supplemented with a 10 minute film of selected test blasts.

Until recently, numeric Film Data Recording has been limited to intermittent type high speed film cameras at speeds up to 500 frames-per-second. A new system has been developed that allows the benefits of numeric film anotation to be incorporated into high speed 16mm and 35mm framing and streak cameras at film speeds up to 11,000 frames-per-second. Externally applied serial data is processed by the system electronics and recorded on each film frame in direct reading numeric form, and in "real time". Principal design objectives were to develop a practical electro-optical-mechanical system that is retrofittable to existing high speed cameras and to assure that the camera's original performance reliability is not compromised by the data recording package.

Since the commercial introduction some twenty years ago of HMI* (Hydragyrum-mercury, Medium, Iodide) type lamps, as a source intended primarily for floodlighting applications, their attraction as a cinematographic light source has been apparent due to their largely desirable characteristics. Use in this field has been restricted due to the absolute requirement for an alternating current supply - with a sine wave supply frame rates are limited to a sub-multiple of the supply frequency with the supply frequency phase locked to the camera frame rate. This has effectively barred metal halide HID lighting from use in high speed photography. The general characteristics of metal halide HID lamps are presented alongside a sample of other light sources. An electronic ballast which has been proven to 12000 Watts in the motion picture industry is then described which overcomes the limitations of the conventional magnetic ballast - the square wave output of the electronic ballast theoretically allows the use of any camera frame rate/shutter angle combination. Finally the suitability of luminaires for high speed photography is discussed.

An intensifier has been developed for use with high speed cine cameras, high speed video and image converter cameras. The Cinemax intensifier gives a real increase in sensitivity of at least 100x for white light events. This can be further enhanced by using wider aperture lenses than are possible with high speed cine cameras. Normally image intensifiers cannot be used in front of high speed cameras because long phosphor decay times cause image stacking. Rapidly changing events become smeared, resulting in poor resolution and low contrast. Cinemax is the first high brilliance intensifier to use sub-microsecond decay phosphors and when placed in front of an image converter camera, speeds up to 1,000,000 f.p.s. are possible. Cinemax is run in continuous mode at low light levels for setting up and in gated mode when recording a high speed sequence. The gated-on period is synchronised with event and camera and is designed to keep the energy dissipated in the final phosphor below damage threshold. Cinemax has successfully been used in conjunction with a Hyspeed cine camera to record the initial burning phase in a lean burn internal combustion engine.

A Conical Shaped Charge (CSC) is an explosive assembly used as a single point penetrator in applications ranging from commercial to national defense. The formation, flight characteristics, and penetration effects of the CSC are of great importance to its users. Because of the extreme velocities of the shaped charge jet, high speed documentation was necessary to research CSC effects. By use of closure switches and initial velocity studies, framing and lighting information for the Cordin 114 Ultra High Speed camera was determined. After the initial data was gathered, it was determined that another source of documentation was necessary. A Hewlett Packard 150 kilovolt flash x-ray system was implemented, enabling more accurate velocity measurements and a better understanding of jet formation and the CSC during flight. Both techniques of documentation were invaluable in the characterization and research of the conical shaped charge.

High-speed instrumentation cameras were required to document the explosively induced structural response of two missile silo simulating structures. The two structures were subjected to simulated airblast pressures in excess of 50,000 lb/in2 and accelerations of 20,000 G's. This extremely harsh environment presented various problems which had to be resolved to permit the intra-structure photography. The high-speed cameras would provide a subjective and quantitative analysis of the structural response. A unique kinematic mounting scheme and shock-isolation system was implemented which ensured the survivability of the high-speed camera equipment.

In this report we will review recent developments in our x-ray streak camera program. The characterization of the linearity and dynamic range of these diagnostics is vital for quantitative analysis of the data, which can then be compared to hydrocode simulations. Examples of time-resolved x-ray imaging and x-ray spectroscopy are presented to illustrate our current capabilities.

We present data characterizing an x-ray streak camera which uses a microchannel plate as a photocathode. The microchannel plate provides good sensitivity to soft x-ray photons and adjustable gain in addition to one dimensional spatial imaging. The streak camera was used in conjunction with a high sensitivity CCD camera and a computer data acquisition system. The resulting detection system has single photon sensitivity, a time resolution of better than 200 psec, and a dynamic range exceeding 103.

The contribution to the degradation of image resolution fidelity in streak camera systems by proximity-focus MCP image intensifiers is explored. These image intensifiers are found to have a significant limitation on overall system performance.

Time-resolved x-ray photography at the University of Rochester's Laboratory for Laser Energetics is utilized to measure the shell trajectories during the implosion of laser fusion targets. Time-resolved measurements of the x-ray emission are made by employing an x-ray streak camera coupled to an x-ray pinhole camera. Streak records are absolutely timed to the incident laser pulse to make comparisons with one-dimensional hydrodynamics computer code simulations. For useful measurements to be made, accurate pointing of the pinhole must be achieved and maintained. We describe methods to characterize the pinhole pointing and show results obtained with this system.