Search

A family of novel devices for electronic imaging is described and several different realizations age presented. Direct electronic Fourier transform (DEFT) cameras employ coupling between surface strain waves, the image intensity, and electrical charge to produce electronic signals representing the spatial Fourier transform of the image at a single pair of terminals. This signal can be easily processed to accomplish pattern recognition, all-electronic image magnification (zoom) and stabilization, motion detection, and focusing. Reproduction of the image is expected to be accomplished by a light valve employing strain-induced modulation of uniform light. Experimental results for light intensity variation in one dimension are reviewed for a silicon chip bulk acoustic wave camera and for a CdS surface wave camera. Some preliminary results are also reported for a fused quartz elastobirefringent light valve image reconstruction device.

Semiautomated tasks frequently require humans to perform highly repetitive boring jobs such as placing objects into machine fixtures. The use of an electromechanical hand design is discussed which may operate in conjunction with industrial robots, part feeders, and minicomputers to perform some of these jobs. Flexibility is achieved with the same hardware by using different control algorithms for differently shaped objects. The design principle which permits simplicity is that the motion which is used to adjust object orientation is also instrumental in the detection of orientation. A ``hand'' was built and a control algorithm to orient a specific object was developed. The control algorithm first recognizes orientation by computing asymmetries and then conditionally adjusts a positional servomechanism in the hand to bring the object to a standard orientation. Tests verified hand performance and indicated restrictions on object shape. Regardless of the initial angle (360° range) about the uncontrolled axis, the computer-controlled hand adjusted this angle to within ±3°. For the prototype, object shape is primarily restricted by the requirement that objects must have their principal axes directed to within ±10° by conventional part feeders and sorters.

Experiments have been carried out that suggest that the observed psychophysical impairment of motion detection in man may not occur equally over the entire visual field. Instead, if impaired motion detection is a consequence of redirection of 'attention' to another portion of the visual field, then movement detection may be impaired the most in the fovea and the least in the region of the subsequent target.

A control stick with a built-in tactile aiding device was tested in a simulated car-following situation. The tactile device gave the driver of a following car information—headway and relative velocity—concerning the state of a lead car. Experimental results (relative velocity and headway variance) with the simulator were compared with those obtained using conventional automobile controls in a similar situation. Sizeable reductions in these quantities, 55 and 85 percent, respectively, were obtained when the tactile display was partially quickened. Some evidence indicated that the driver behaved as an amplifier when using such a display.

ASW effectiveness is strongly dependent on whether sonar contact can be maintained long enough for a attack to be accomplished. Proper design and use of weapons systems both depend upon the probable duration of contact maintenance. A number of stochastic models of propagation and detection variability are considered. A method is described of fast-time tactical simulation which makes use of such models to evaluate the overall effectiveness of the sonar/weapon system combination.

Spectral and polarization sensitivity measurements were made at several levels (retina, first and third optic ganglion, cervical connective, behavior) of the dipteran visual nervous system. At all levels, it was possible to reveal contributions from the retinular cell subsystem cells 1 to 6 or the retinular cell subsystem cells 7 and 8 or both. Only retinular cells 1 to 6 were directly studied, and all possessed the same spectral sensitivity characterized by two approximately equal sensitivity peaks at 350 and 480 nm. All units of both the sustaining and on-off variety in the first optic ganglion exhibited the same spectral sensitivity as that of retinular cells 1 to 6. It was possible to demonstrate for motion detection and optomotor responses two different spectral sensitivities depending upon the spatial wavelength of the stimulus. For long spatial wavelengths, the spectral sensitivity agreed with retinular cells 1 to 6; however, the spectral sensitivity at short spatial wavelengths was characterized by a single peak at 465 nm reflecting contributions from the (7, 8) subsystem. Although the two subsystems exhibited different spectral sensitivities, the difference was small and no indication of color discrimination mechanisms was observed. Although all retinular cells 1 to 6 exhibited a preferred polarization plane, sustaining and on-off units did not. Likewise, motion detection and optomotor responses were insensitive to the polarization plane for long spatial wavelength stimuli; however, sensitivity to select polarization planes was observed for short spatial wavelengths.

In this paper a survey is presented of the theoretical and experimental data obtained in the last eight years about both the shape of the pulses supplied by semiconductor detectors and the techniques for obtaining from these pulses a fast timing information. Moreover the problem is considered of extracting from the shape of the pulse some information on the type or the impact angle of the detected radiation.

The nonlinear system identification technique through white-noise stimulation is extended to multi-input, -output systems with consideration given to applications in the functional study of the nervous system. The applicability of the method is discussed in general and in particular for the motion detection neuronal system of the fly. Two series of experiments are performed; one with moving striped-pattern stimuli and the other with spot stimuli of fluctuating intensity. In both cases nonlinear dynamic models are derived which describe the system with considerable accuracy over the frequency range of 0.2–50 Hz and a dynamic amplitude range of about 40-1. These models are able to predict accurately all the discrete experiments so far performed on this system for which the models are applicable. The differences in dynamic characteristics between the corresponding system of the Musca and Phoenicia families of flies are minor except for a difference in latencies and if the difference in geometry of their faceted eyes is taken into account. The large field response of the motion detection unit is a linear weighted summation of all the smaller field highly nonlinear subsystems of which the large field is comprised.

Several classes of interneurons in the optic lobes and brain of the insects, Musca domestica and Calliphora phaenicia, have been studied in detail. Visual stimuli have been categorized on the basis of the properties of intensity, form, and motion. Response characteristics of the classes of neural units are described with respect to these three classes of visual stimuli. While those units that detect motion in select directions have a tonic response, form detection units have a phasic response only. Through correlation of the responses of these classes with visual stimuli, it is shown that these units integrate the responses of other units which have very small visual fields. The small-field units are presumed to integrate the output of a small group of adjacent retinula cells and to respond differentially to intensity, form, and motion. It is shown that the response of both form and motion detection units is independent of the direction of pattern intensity gradation. As a consequence of this independence, it is further shown that failure to detect motion properly must start at a spatial wavelength four times the effective sampling station spacing rather than twice as has been predicted previously.

Television signals contain a great deal of frame-to-frame redundancy because picture areas are scanned in every frame whether they have changed or not. That portion of the signal describing stationary images need not be retransmitted in every frame if adequate memory is provided at the receiver. The signal describing moving images must be transmitted, of course, but it requires progressively less fidelity as the motion increases. Several techniques are described for obtaining efficient transmission by taking into account the similarity in the signal from frame to frame. Basic to most of these techniques is the need to separate the signal into segments that have changed significantly since the previous frame and ones that have not changed.

High-resolution position detection using a one-dimensional zero-dead-band photosensor receptor has been shown [1] to be possible by applying a high-frequency triangular perturbation to a normalized target image. It is shown here that for a nonzero dead-band ? between the receptor elements, position detection becomes increasingly uncertain as ? increases. By periodically changing the magnitude of the perturbation and the system gain, it is shown that accurate position detection is still possible for a receptor with nonzero dead-band.

An approach to the analysis of threshold learning suggested by the classical theory of Brownian motion is presented. In particular, it is shown how a nonlinear Langevin equation represents the motion of the threshold of a trainable signal detector, and how a Fokker-Planck diffusion equation yields an estimate of the shape of the probability density of the threshold. Our results are applicable to all trainable signal detectors in which the training procedure 1) raises the threshold in response to a false alarm, lowers the threshold in response to a false rest, and keeps the threshold unchanged in response to a correct decision, and 2) adjusts the size of the threshold increment by an amount that depends only on the trial number, and such that the threshold can eventually reach any real number.

Nervenfasern, die die optischen Ganglien beider Kopfseiten von Musca und Phaenicia verbinden und die selektiv auf Musterbewegungen antworten, zeigen binoculare Wechselwirkungen. Details dieser Wechselwirkungen und der Ort an dem sie in den optischen Ganglien stattfinden wurden ermittelt. Die Bedeutung dieser Einheiten fur die direkte Steuerung von Wendereaktionen beim Flug wurde ebenfalls untersucht.

THE PHOTO- OPTICAL SYSTEMS OF MARINE ANIMALSBy Kiyoe MizusawaChrysler Corporation Space DivisionAbstractIn a previous paper by the author(Ref. i$) it was concluded that marineanimals primarily use color and motionas cues for visual performance almostexclusively, in contrast to land ani-mals who use shape, size and distancestimuli as cues, although they havestructurally similar eyes. This paperexplains to what extent marine animalsuse color and motion stimuli as op-posed to shape, size and distance ascues in marine environment. In thepaper the behavioral measurementsmethod was emphasized.IntroductionThis paper is an extension of apaper presented by the author at theSeminar of "Underwater Photo -Optics,"sponsored by the U. S. Naval MissileCenter Pt. Mugu and the Society ofPhoto -optical Instrumentation Engi-neers (Ref. l8). In the previouspaper, it was concluded that marineanimals primarily use color and motionas cues for visual performance almostexclusively in contrast to land ani-mals who use shape, size and distancestimuli as cues, although they havestructurally similar eyes. This paperexplains to what extent marine animalsuse color and motion stimuli as opposedto shape, size and distance as cues inmarine environment.The first objective of the pre-vious paper was to interpret the cap-abilities and limitations of man'svisual systems underwater in relationto fish based upon phylogenetic com

The properties of movement detectorsin man, inferred from psychophysical tests, are remarkably similar to those found in other animals such as the frog, pigeon, rat, rabbit, cat, and monkey. The smallest fields in man are about 2° in diam., increasing to about 20° in the periphery. The fields are directionally specific, and exhibit surround inhibition. Preferred directions for centers are horizontal and vertical; the inhibitory surround is not directionally selective. Optimal sensitivity to speed for foveal objects occurs at about 3 to 6°/sec, and the contralateral pathway is about 1.4 X more sensitive to motion than the ipsilateral route. The inferred properties of motion detectors appear to represent the behavior of a subcortical system, presumably involving the superior colliculus or pretectum. Under certain conditions, this movement detection system may be shown to interact with stereoscopic mechanisms that are believed to be cortical.

The multi-aperture image dissector has the ability to give simultaneous information on the brightness of spatially separated points of an optical image, due to the multiplicity of dissecting apertures. Several applications for this multiple readout concept have been proposed and demonstrated. Details of systems for edge detection, tracking and use in photogrammetry, incorporating the multi-aperture image dissector, will be given. Possible motion detection and shape recognition schemes will be outlined. A low light level development of the multi-aperture image dissector will be discussed.

Because of the accuracy needed in the operation of photo-optical systems of simulation, the present paper attempts to measure the differences in sensitivity of motion perception in the various portions of the retina. There is a noticeable difference between the central and peripheral retinal areas in the detection of angular movement and in threshold excitation. The peripheral area dominates the central area in motion detection because it has more rods, thus having a lower threshold excitation and a lower threshold of critical fusion frequency (c.f.f.). By measuring the c.f.f. gradient at varying visual angles on the nasal, temporal, inferior and superior coordinates, it was found that the peripheral area, specifically temporal area, has lower threshold to' motion perception than the central area.© (1969) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

With the advancement of long duration of manned space flight and prolonged submergence of the nuclear powered submarine, research on the effects of atmospheric contaminants on human visual performance has been increased. Atmospheric ions could cause changes in efficiency of visual performance. The present paper is concerned with systematic review of articles on the effect of atmospheric ions on visual performances. In general, exposure to certain amount of negative ions enhance certain aspects of visual parameters such as brightness discrimination and motion detection, while certain amounts of positive ions deteriorate most visual performances. The need for the control of air ions in maintaining optimal environmental conditions has been suggested in this paper.© (1970) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

In normal vision, eye movements cause the image to move on the retina. Special apparatus can stabilize the image on the retina so it cannot move, and vision fades away. Previous methods for stabilizing the image were either optical systems or complex computer-controlled feedback systems. A simpler feedback system is presented which detects eye motion and rotates a mirror through which the target is viewed, to exactly compensate for eye motion. Unlike previous optical systems, this new system has no lenses in either the viewing or the image-forming path to limit the spatial resolution.

The basic principle of motion detection by fibers of the optic lobes of flies were studied with a pair of small spots and a variety of paired intensity variations. These show that the process of correlation of adjacent field regions to detect motion is confined to a small area. The presence of small field units with small field adjacent inhibition in the system was detected. The optimum spot spacing for maximum reactions corresponded to the facet spacings. Selective motion detection responses from minimum information consisting of evaluating the difference between the spot intensities and the rate of change of the trailing spot relative to the motion direction was shown. However, additional properties best determined by white-noise experiments designed from this study were found.

A device has been designed and constructed to aid a blind radio station operator to cue phonograph recordings for broadcast. The device uses a hand-held detector to scan the record surface and generates a tone that changes markedly in frequency when the detector is passed over a band between selections. The use of the device allows a blind operator to work unassisted by sighted persons.

This letter describes a technique for detecting motion due to self-heating of small semiconductor devices. The technique utilizes holographic interferometry coupled with holographic microscopy. The optical arrangement described is similar to a normal illumination microscope. Although oblique illumination of the object could be used, the former technique facilitates the observance of the fringe pattern, indicative of motion of the device.

Motion thresholds were determined for the fovea and peripheral retina with and without correction for peripheral refractive error. With correction, motion thresholds decreased and inidividual differences disappeared. These results imply that under normal observation conditions, peripheral sensitivity is limited mainly by dioptric rather than retinal variables.

Jamming experiments were conducted to assess the effectiveness of several different types of signals for interfering with the ability of echolocating bats (Eptesicus fuscus) to discriminate target range. Random interfering noise covering the bat's sonar bandwidth of 25–50 kHz must be sufficiently intense for the echo signal‐to‐noise ratios to approach 0 dB before marked breakdown of echolocation occurs. This is true for continuous noise, intermittent noise, and for noise incorporating internal correlation at the same period as the travel time of sounds going to and from the targets. FM interference is 30 to 40 dB more effective in jamming the bat than random noise of the same spectrum. The results suggest that the bat's sonar receiver applies frequency‐sweep criteria to incoming sounds to decide whether potential echoes are present. Neurons with properties appropriate to this filtering function, analogous to motion detection in vision, are known to exist in the bat's auditory nervous system.