The distance one can see using an underwater lamp is limited by veiling luminance from the lamp's beam. The simplest and least expensive way to reduce this luminance is geometrical. One separates the lamp as far as possible from the imaging aperture (eye or camera) and looks through as little of the light beam as possible. For a downlooking geometry, e.g., searching the sea floor, a geometry that directly illuminates only the extreme lateral edges of the field of view and lets scattered light suffice to illuminate the interior of the field is investigated. Image intensity and veiling luminance were calculated as functions of distance for two geometries assuming typical numbers for scattering and absorption of light in the sea. Regarding theoretical factors presented here, indirect illumination is competitive with conventional direct illumination. Regarding practical factors such as quantum-limited image detectors for extended range, the indirect scheme has merit in keeping excess light out of the detector.

An analytical expression for the underwater radiance distribution due to a purely 'delta function' sun is discussed. The expression derives from a WKB evaluation of the path integral solution for time-dependent radiative transfer, integrated over long times, and does not involve a small-angle approximation. However, a diffusion-limiting length scale previously found in the small-angle approximation also arises in this evaluation, suggesting that is plays a physically important role (independent of approximation schemes) in governing the structure and evolution of the radiance distribution. In its present form, the analytical expression reproduces the shape of the downwelling radiance distribution for angles as large as 90 degree(s), but is inadequate for the upwelling component. However, the poor modeling of the upwelling component is not a limitation of the WKB approximation, but is most likely due to the simplistic treatment of the phase function. An effort is underway to more carefully handle the phase function within this WKB framework.

A range-gated underwater imaging system is described that utilizes a frequency doubled, Q- switched Nd:YAG 30 Hz laser source for illumination and a fast-gated microchannel plate intensified CCD array camera for detection. Laser pulse widths of 7 ns are timed relative to comparable camera gates with subnanosecond jitter. Results for a test bed system obtained in a 4 X 4 X 40 ft water tank with various targets are presented. Water quality was varied with the addition of progressive concentrations of Maalox as scattering agent, and monitored with a home-built laser transmissometer. The 1/4 in., high contrast target lines were able to be resolved at ranges exceeding four attenuation lengths. A simple analytical model for image signal-to-noise ratios is presented and a straightforward polarization discrimination scheme suggested for contrast enhancement. Polarization optics were incorporated into the range-gated test bed system and results obtained for targets of varying characteristic depolarization. In all but a few cases where target and background depolarizations were similar, the signal to noise is enhanced in spite of rejection of orthogonal polarization signal. Issues regarding the evolution of the test bed system to field operation are discussed and significant progress in the development of appropriate miniaturized and ruggedized components is presented

The polarization state of light in the ocean can be used to enhance visibility. The consequences of scattering from nonspherically-symmetric particles on light propagation and visibility in the ocean was investigated. To calculate scattering from nonspherical marine microorganisms, it is usually necessary to resort to approximate methods. One promising approximation is the coupled-dipole approach in which an arbitrarily-shaped object is divided into a number of identical elements arranged on a cubic lattice. Each element is treated as a spherical, dipolar oscillator with its polarizability specified by the real and imaginary parts of the index of refraction. Interactions between dipoles are included by determining the field at a particular dipole due to the incident field and the fields induced by the other dipole oscillators. The scattered field is then the sum of the fields due to each oscillator. The coupled-dipole method is promising because, in principle, an organism of any shape can be modeled, and all 16 elements of the scattering matrix calculated. This approach has been applied to calculate scattering from spherical particles to verify the limits of the approximation, and from other shapes to investigate the effects of nonsphericity and chirality on scattering. In particular, all 16 Mueller matrix elements for the scattering were calculated from a finite cylinder, a single- strand helix, 14-strand helix, and ensembles of these particles. The effects of pitch, size, wavelength, and complex index of refraction were investigated. The results provide insights into the magnitude and type of depolarization effects associated with various marine microorganisms containing these structures.

Light scattering measurements were performed on single, immobilized dinoflagellates as well as on suspensions of the species Prorocentrum micans. The normalized Mueller scattering matrix element S₁₄, which indicates an ability to depolarize circularly polarized light, is reported for both cases. The measurements involving single cells were performed on Prorocentrum micans, Gonyaulax polyhedra, and Crypthecodinium cohnii. The results show that the previously reported large S₁₄ signal is not peculiar to P. micans. Time-dependent measurements of live cultures of P. micans show a large, high-frequency S₁₄ signal. This signal is a diurnal function of the time of day, with a maximum at midnight. Investigations of the relationship between single cell and suspension measurements reveal that the large angle- dependent S₁₄ peaks from immobilized single dinoflagellates are responsible for a large time-dependent S₁₄ signal at 90 degree(s) in suspension measurements. The results of these experiments provide further evidence for the hypothesis that the chromosomes of the dinoflagellates are responsible for the large observed S₁₄ signals. The unusual depolarization properties of dinoflagellates should be considered when using polarized light to enhance image contrast in underwater imaging.

A system is described that produces high quality images through turbid waters by means of time encoded reflected light transmitted by scattering. The system consists of a compact battery operated laser scanning unit that scans the underwater scene with the laser beam in a manner similar to a television raster. Light reflected from any object in the scene varies in accordance with the reflectance of the minute spot being illuminated. This time varying intensity (TVI) signal is transmitted through the water to a remote receiver by both scattered and unscattered light where the received signal may be stored and/or displayed. The underwater laser scanning unit can be moved freely about the field of interest by scuba diver or ROV, unencumbered by entangling umbilicals, and can send real-time images over distances of 15 to 20 attenuation lengths to observers in a shirt-sleeve environment for critical viewing on an image display monitor. This previously undescribed system was developed in the early 1970s for proof of concept tests and used technology that is now 18 or more years old. The physical principles and the experimental hardware are described and examples are given of images providing exquisite detail that were made in an experimental tank together with some images obtained in ocean trials.

The objective of this paper is to utilize Wells' results to compute the medium decay function, D(Γ), from which is easily determined the germane MTF for a stated range, R) using both the value of b computed with the above mentioned relationships and experimentally determined values of b. The resulting decay function expressions will be compared to provide a measure of the errors introduced by using the various estimation methods. Finally, comments will be made on which model appears to provide the most accurate estimation of b for different water mass types.

Measurements of the point spread function (PSF) were made in the Sargasso Sea in December, 1989. The water column during this period was extremely homogeneous therefore this data set became an obvious candidate to test the theories of Wells. This paper describes the results of this study. It's shown that the formalism (hence the implicit small angle approximation) works reasonably well for this case, with an average difference of 12% between the measured and predicted PSFs.

The effect of a detector on the ocean radiance field being measured has been investigated -- first, in a simple model that accounts for perturbations in the radiance field introduced by the detector shadow and by light reflecting from the detector, and second, by a complete three- dimensional solution of the radiation transfer equation that calculates the perturbations by using a computer model of a 4(pi) radiance detection system (RADS²) with sky and ocean parameter choices to match California coastal water. Results show that the presence of the detector has little influence on downwelling measurements nor do changes in detector reflectance. There is a substantial change due to the detector of upwelling radiance values. This may cause measured values to be reduced by as much as 30% below the unperturbed radiance for systems of the size of RADS (0.5 m) for light coming from the region of the detector shadow. The upwelling irradiance for this case is reduced 10%. Results are given for sensor 0.5 m high and having diameters from 0.125 m to 2 m.

This paper describes a novel experiment to simultaneously measure, over relatively long, near- vertical paths in unmixed ocean water, the spectral absorption and attenuation coefficients and the volume scattering function from which the scattering coefficient can be computed by integration. The experiment can therefore be used to test the widely used closure property in ocean optics. Another important aspect of these measurements is that they are indicative of the undisturbed, bulk inherent properties of the ocean water as opposed to conventional, short- path measurements, which appreciably disturb the measured water volume. This experiment was recently carried out during ONR's ocean optics cruise in August 1990 in Pacific waters off the coast of southern California.

A computer model to simulate the formation of color underwater images has been developed. The model simulates the appearance of an underwater image as seen by an underwater camera. The images are intended to portray the appearance of a planar reflectance map that has been illuminated with artificial illumination. The input to the model consists of the geometry of the camera, artificial lights, the environmental constants that govern the propagation of light underwater, and the reflectance values of the map. To simulate the appearance of color images the spectrally varying nature of these inputs has been taken into account. This article describes the algorithm and illustrates the images that can be obtained as a function of different water types and camera/light configurations. A specific imaging geometry is used to simulate the appearance of objects that would be seen by a towed underwater imaging platform. The images have been synthesized in two types of ocean water at several distances. The results indicate the important role that frequency dependent scatter and absorption play in the process of underwater image formation.

Submerged spectral reflectance measurements made on paint samples using two different techniques are compared. With the first technique spectral measurements are made with a simple thin water film measurement technique originally used for photopic viewing. A comparison of these measurements with a second technique in which the measured sample is immersed in water in a cylindrical container and the submerged reflectance is measured with a goniophotometer shows good agreement for wavelengths from 420 to 700 nm.

The design and construction of an isotropic light source is described. The instrument's design is well suited to underwater applications. Any combination of light sources (flashlamps, CW sources, or lasers) and spectral filters (absorption or interference) can be simultaneously incorporated to provide a single isotropic source with any desired temporal and spectral characteristics. This design was recently used in an underwater experiment that required a synchronously triggered isotropic flash to simultaneously measure absorption, scattering, and attenuation. The design of this particular isotropic source is presented along with data demonstrating the isotropy of the light field produced by the source.

An imaging model for predicting the performance of underwater range-gated imaging systems has been developed. Using this model, image quality as a function of the inherent optical properties of the water and as a function of the parameters of the optical system (e.g., laser pulse width, camera gate width and time, laser and camera FOVs, and laser-camera separation) can be determined.

The in situ imaging of marine particles shares many of the signal-to-noise difficulties of underwater imaging in general. Natural and traditional artificial illumination, for example, allow light scattered from particles outside the imaging volume, reducing the image contrast. Sizing and classification of small-particle images (magnification approaching 1 or more) have additional difficulties associated with a limited depth-of-field and the resulting noise from illuminated but unfocused targets in the field of view. Moreover, target sizing and classification are uncertain without individual target range information. The new marine particle imaging instrument to be discussed employs diode laser illumination (675 nm) with line-generator optics to produce a thin light sheet at the system focal plane. This light sheet and narrow-band, optical filters are utilized to minimize noise associated with diffuse ambient illumination since significant red ambient illumination is lost below 5 m depth. It also removes the uncertainty involved in the determination of the three-dimensional position and size of a target in a two-dimensional image. An additional problem inherent in marine particle research is that the size of the particles of interest ranges over several orders of magnitude (micrometers to centimeters diameter). The instrument addresses this problem of scale with coincident video imaging systems of high and low spatial resolution. Shape-generated feature vectors and particle optical attributes are extracted from digitized particle images and utilized in an automatic particle classification scheme. The strategy is multidimensional and incorporates a pattern recognition algorithm rooted in the theory of moment invariants.

A backscatter sensor has been developed for rapidly measuring, in situ, the volume scattering function (VSF) in the backward direction. The backscatter sensor uses a bistatic optical geometry to measure backscatter from a small volume of seawater over a range of scattering angles from approximately 115 degree(s) to 170 degree(s). The calibration of the sensor yields a weighted, angular averaged value of the VSF with a centroid located at a scattering angle of about 150 degree(s). The backscatter sensor design is based on a sensitive synchronous detector and pulsed, light-emitting diode that has been used at visible and near-infrared wavelengths. The entire sensor package, which includes circuitry for digitizing the signal, is contained in a compact, rugged housing. The sensor has been deployed both in towed arrays and in stationary profiling mode. Scattering profiles from two recent deployments are presented.

A previously proposed phenomenological model of the beam spread function in seawater is extended to include dependence on the asymmetry parameter of the scattering phase function.

The required dynamic range of a practical undersea laser-based sensor is limited by its range- to-target. The dynamic range of the sensor needs to accommodate the variations in target reflectivity and specularity, water scattering and absorption, and mode of operation of the sensor at the target range. For example, the same sensor may be used to sense a diffuse flat black target and a shiny metallic mirror at normal incidence. Similarly, the same sensor may be deployed both in murky coastal waters and in clear deep oceanic waters. Since a detector used in a sensor has a limited (intrascene) dynamic range of perhaps 40 dB, methods of mitigating the required dynamic range at the sensor must be employed. Further, the required times scales of each specific dynamic range absorber must be considered. Therefore, a systematic optical budget approach considering time scales is employed. After considering maximum laser power out and minimum detectable power, the remaining available dynamic range is allocated to specific devices based on time responses matched to requirements. Devices are discussed and classified according to their parameters, and a final recommended system design is presented. The final system consist of a continuously variable neutral density filter wheel and a galvanometrically scanned variable neutral density filter wheel. Other devices and their merits are discussed.

When photogrammetry is used to research the moving target in water, one can encounter with two conditions. One of them is underwater photogrammetry, another is the photogrammetry through the big water-pot made from transmitting glass. In both cases, there is a sight glass between the taken target and camera. This glass will have an effect on the results of photogrammetry. The problems of the quality and accuracy of photogrammetry under this condition are discussed.

Subsea laser radar has a potential for accurate 3-D imaging in water. A prototype system has been developed at Seatex A/S in Norway as a prestudy for the design of an underwater laser radar scanning system. Parallel to the experimental studies, a numerical radiometric model has been developed as an aid in the system design. This model simulates a raster scanning laser radar system for in-water use. Thus this parametric model allows for analysis and predictions of the performance of such a sensor system. Experiments have been conducted to test a prototype laser radar system. The experimental system tested uses a Q-switched, frequency doubled, Nd:YAG solid state laser operating at a wavelength of 532 nm, which is close to optimal for use in water due to the small light attenuation around this wavelength in seawater. The laser has an energy output of 6 (mu) J per pulse 1 kHz pulse repetition frequency (PRF) and the receiver aperture is approximately 17 cm². The laser radar prototype was mounted onto an accurate pan and tilt unit in order to test the 3-D imaging capabilities. The ultimate goal of the development is to provide an optical 3-D imaging tool for distances comparable to high frequency sonars with a range capability of approximately 30 - 50 m. The results from these experiments are presented. The present implementation of the scanning laser radar model is described and some outputs from the simulation are shown.

The design of two advanced underwater lenses for use with high-resolution, charged coupled device (CCD) still cameras is presented. Some practical aspects of lenses for CCD cameras are discussed and how the customer's requirements led to different choices of water-lens interface is demonstrated.