Current Laser Line Scanner (LLS) sensor performance is limited in turbid water and in bright solar background conditions. In turbid water, backscattered and small angle forward scattered light reaching the receiver decreases underwater target contrast and resolution. Scattered solar energy reaching the detector also decreases detection sensitivity by increasing receiver noise. Thus, a technique which rejects unwanted, scattered light while retaining image-bearing photons is needed to improve underwater object detection and identification. The approach which we are investigating is the application of radar modulation and detection techniques to the LLS. This configuration will enable us to use optical modulation to discriminate against scattered light. A nonscanning mock-up of an existing LLS, the Electro-Optic Identification sensor, has been developed with off-the-shelf components. An electro-optic modulator will be added to this system to create a modulated LLS prototype. Laboratory tank experiments will be conducted to evaluate the performance of the modulated LLS as a function of water clarity and solar background levels. The new system will be compared to its unmodulated counterpart in terms of target contrast.

A Streak Tube Imaging Lidar (STIL) system has recently demonstrated high resolution 3D imaging in ocean water. Lateral resolution of better than 1/2 inch has been obtained through 36 feet of ocean water, with range resolutions of better than 1 inch. The resulting 3D data may be processed for imaging of floating, moored, and bottom objects. Processed results include both conventional contrast maps of bottom targets, as well as range maps which represent the height of the object above the surrounding bottom. Data analysis and modeling was performed to process and interpret the resulting high resolution 3D images. This paper describes the concept, design, implementation, and performance of STIL for ocean imaging, and discusses applications of this high resolution 3D imaging capability to ocean surveillance.

Lite Cycles has developed a new type of range-gated, LIDAR sensing element based on Raman image amplification in a solid-state optical crystal. Marine Raman Image Amplification (MARIA) is a feasible technology for producing high-resolution imagery in an underwater environment. MARIA is capable of amplifying low-level optical images with gains up to 10⁶ with the addition of only quantum-limited noise. The high gains available from MARIA can compensate for low quantum efficiency detectors. The range-gate of MARIA is controlled by the pulsewidth of the amplifier pump laser and can be made as short as 30 - 100 cm, using pump pulses of 2 - 6.7 nsec FWHM. The use of MARIA in an imaging LIDAR system has been shown to result in higher SNR images throughout a broad range of incident light levels, in contrast to the increasing noise factor occurring with reduced gain in ICCDs. The imaging resolution of MARIA in the marine environment can be superior to images produced by a laser line scan or standard range-gated imaging system. MARIA is also superior in rejecting unwanted sunlight background, further increasing the SNR of images. MARIA has the potential of providing the best overall system resolution and SNR, making it ideal for the identification of mine-like objects, even in bright sunlight conditions.

The Naval Oceanographic Office (NAVOCEANO) and Planning Systems INcorporated are developing the Littoral Optics Geospatial Integrated Capability (LOGIC). LOGIC supports NAVOCEANO's directive to assess the impact of the environment on Fleet systems in areas of operational interest. LOGIC is based in the Geographic Information System (GIS) ARC/INFO and offers a method to view and manipulate optics and ancillary data to support emerging Fleet lidar systems. LOGIC serves as a processing (as required) and quality-checking mechanism for data entering NAVOCEANO's Data Warehouse and handles both remotely sensed and in-water data. LOGIC provides a link between these data and the GIS-based Graphical User Interface, allowing the user to select data manipulation routines and/or system support products. The results of individual modules are displayed via the GIS to provide such products as lidar system performance, laser penetration depth, and asset vulnerability from a lidar threat. LOGIC is being developed for integration into other NAVOCEANO programs, most notably for Comprehensive Environmental Assessment System, an established tool supporting sonar-based systems. The prototype for LOGIC was developed for the Yellow Sea, focusing on a diver visibility support product.

Extensive measurements were carried out of the optical properties of seawater off the East and West Coasts of Canada using a full-spectrum near-forward angle nephelometer (NEARSCAT). Using a new phase function adapted specifically to scattering from water borne particles, we analyze the data from coastal waters in the Straits of Juan de Fuca and in the Gulf of St. Lawrence. We show explicitly how the spectral properties can be combined with the angular properties to more reliably extract the Junge exponent of the particle size distribution and the mean index of refraction of the particles. We obtain a simple analytic expression for the normalized cumulative phase function that can be used to compute the backscatter ratio, and its explicit wavelength dependence. Accurate estimation of this wavelength dependence is required for accurate hyperspectral image prediction.

Multispectral and hyperspectral sensors are being used for remote sensing and imaging of ocean waters. Many applications require the compression of hyperspectral data to achieve real-time transmission or exploitation. Hyperspectral data compression or reduction has been accomplished using techniques based upon principal component analysis or linear unmixing. Alternatively, data compression (reduction) may be performed by band selection, or band selection may be preliminary to either of the other compression techniques. Band selection also has implications for sensor design and the stability of estimates of processing parameters. In this study, we address the question of which bands are the most efficacious for imaging submerged objects, such as whales, using an anomaly detector, or a matched filter. Bands are selected by optimizing a detection criterion subject to a constraint on the number of bands. The technique is applied to give hyperspectral data sets, and the optimum bandwidths and centers are determined. The loss in performance from selecting reduced numbers of bands is tabulated and the need for adaptively selecting reduced numbers of bands is demonstrated.

We have recently developed a model of the optics of the underwater environment that includes the effects of scattering and absorption. The model is used to predict the image degradation that occurs in water. The model is applicable to standard light source illumination, range- gated illumination and laser line scan illumination. The model takes as input standard pictures and produces an output degraded images with the effect of scattering, absorption and detector noise added in. The model allows us to test against realistic images the performance of various image recovery and enhancement techniques and to compare the various active imaging systems that could be used in underwater identification systems. Using a previously developed sea surface and sky irradiance model developed at DREV, we are currently extending the model to handle underwater imaging from airborne platforms by a multi-pulse laser bathymetry system and by a range gated laser-imaging system.

Under sponsorship from ONR, Planning Systems Incorporated is developing the Generic Lidar Model (GLM) to compare different Lidar systems (pulsed/CW, airborne/in-water, monostatic/bistatic, wide/narrow beamwidth) using a consistent treatment of the environment. The model uses Modulation Transfer Functions to incorporate environmental influences on image detection and blurring. Environmental parameters include absorption, the depth-dependent effect of a scattering layer within the atmosphere and the water column, backscattering, bottom and target reflectivity, wind-driven air/sea interface changes. System parameters considered include source/receiver geometry, various spot and scan configurations, optical defocusing, and power. System electronics and image processing are being incorporated. GLM presently models a pulsed, range-gated system and a CW laser line scanner with the help of interactive GUIs which allow the user to instantly see and evaluate the effect of changes in important system and environment variables. Optical profiles obtained during the Oceanside '95 exercise sponsored by NRL's Littoral Optical Environment program were used to test GLM predictions. During this exercise, the optical signatures of passing internal waves were noted as having a dramatic effect on a laser-line scanner deployed. Profiles from the exercise were used to predict laser-line scanner performance before, during, and after the passage of the internal waves.

In this first of a series of two papers, previous work is extended to include analysis of the effect on various target recognition algorithms of error reduction techniques that partially compensate interfacial topography mis-estimation. Additional error sources include (1) accrual of spatial error in the image restoration process due to blur resulting primarily from optical effects within the water column, and (2) uncertainty in sensor platform motion. The impact of such perturbations on the accuracy of interframe coregistration, which is essential for multiple-view analysis, is also discussed. A companion paper describes algorithms that pertain to target feature detection and partial restoration.

Reconstruction of submerged targets from multiple views (called multi-look imaging or MLI) is a technology adapted from machine vision and tomographic imaging whose utility is emerging in airborne battlefield surveillance applications. In summary, MLI integrates a sufficient number of views that may have different portions of a target or scene visible. The integration process is designed to support semi- automatic derivation of a solid model that portrays an imaged target or scene. This approach is advantageous for target and feature location, especially in the presence of partial obscuration due to turbid optical media, overlying cover, or occlusion by other target objects.

Our previous comparisons of the SNR and resolving power of various lidar approaches used modulation transfer functions to estimate the effects of surface waves. That technique yields statistical performance estimates, but gives little insight into the appearance of individual images. This paper presents a complementary method, simulating images for the various systems. The images are generated with a numerical code combining ray tracing and small-angle scattering theory with a state-of-the-art, dynamic, surface-wave generator. The resulting images realistically simulate the refractive effects of the ocean waves, providing visual confirmations of our previous analytical results. They demonstrate the excellent contrast of a STIL for high-resolution image classification. However, the superior energy utilization of the LRG approach, makes it a better method for airborne lidar imaging, with an advantage that grows exponentially with depth. Likewise, a compact rectangular array of time- resolved pixels perform search tasks better than does a STIL. These simulations will also allow us to develop optimal strategies for image-by-image processing of LIDAR data.

In recent years, owing to a lack of field experiment programs, the R&D in lidar remote sensing of natural waters in Russia is restricted almost entirely to laboratory development of novel elements for airborne systems and theoretical investigations. In this report a theoretical approach based on statistical formulation of signal detection problem is shown to provide an effective tool for comparison of different lidars (and separate elements) effectiveness for various applications--bathymetry, water column sounding, submerged objects detection, etc. The approach quite often leads to unexpected answers to `obvious' questions like Do one really needs a large input pupil for lidar receiver? What is optimal filter bandwidth and angle aperture? What is the best wavelength for oceanographic lidar?

On the base of a newly developed theory of image formation, which includes the effect of shadowing of the space behind an illuminated and observed object, the receiver time gating of an airborne ocean lidar system is simulated. Depending on the timing of the start of the gate, a reflected target or its shadow may be seen. The contrast and signal-to-noise ratio (SNR), for an image of the directly observed object (reflection) and the image of its shadow (obscuration), are compared. It is shown that in many cases the SNR of the object in the `shadow' (obscuration) mode can be greater than the SNR of the object observed in reflection compared at the same depth. In addition, the obscuration mode is advantageous for improving system performance in cases involving a priori choices of the start and duration of the gate. These advantages are most pronounced while observing an object in turbid shallow water through a windy roughened sea surface.

We propose a method of numerical simulation of an image formation in vision systems looking through the waved water surface. The simulation of radiation propagation through the waved sea surface and sea depth is realized in ray approximation. Waviness is simulated as evolving in the time random surface with given spectrum of elevations and dispersion relation. New values of angular, energy, and polarization parameters of each ray in the beam after its refraction and reflection at a given surface point are calculated by means of Fresnel formulae. Seawater medium is split in the layers with thickness chosen from the condition that water characteristics can be considered as constant and that multiple scattering of a ray can be neglected within a layer. Attenuation of beam energy corresponds to the length of the beam path in the layer, and probability of beam scattering at a random angle within a layer is determined by phase scattering function. We have obtained numerically instantaneous and averaged over different exposition times images of 2D self-luminous and located mira made up from spokes and situated at various depths for various values of wind velocities and seawater scattering coefficients. Influence of water and waviness parameters on the image quality has been analyzed.

Many single shot, single pixel underwater LIDAR systems employ high performance photo-multiplier tubes (PMTs) in their receivers. While PMTs offer high gain with a low noise factor, the dynamic range of the output signal is limited by signal induced artifacts. These artifacts include decaying signals with long time constants and short duration `ghost reflection' signals. This paper will characterize the signal-induced artifacts for various single pixel PMTs with different photo-cathode materials and dynode materials.

An introduction and rationale for the papers presented in this session follows.

Experiments with a blue-green laser radar system were conducted off the coasts of Ireland and Scotland in June, 1999. The purpose of this test was to measure the effect of the water optical properties on the polarization state and decay rate of the lidar return signal. The lidar system, the K-meter Survey System (KSS), was configured to transmit linearly polarized light and to receive cross-polarized light in one channel and both polarization in the other channel. Several oceanographic ground truth instruments were used to measure the water optical properties, including transmission, absorption, backscatter coefficient, diffuse attenuation, temperature and salinity, as a function of depth. The KSS was mounted on the bow of one of the UK survey vessels, the HMS Roebuck, and the oceanographic instruments were deployed with a deck-mounted winch. The results presented in this paper were obtained both inside and outside of the continental shelf. Since these regions were characterized by different water optical properties, the sensitivity of the lidar return signal in terms of decay rate and polarization to different water clarities was determined.

The Fanbeam Spectral Imager (FSI) system is under development as a tool for mine countermeasures, however, the system has potential for a variety of additional underwater remote sensing needs. Characteristics common to each intended application include high area coverage rate acquisition and 3D spatial by spectral data. The FSI system has been designed around a modular approach which utilizes separated illumination and detection modules to achieve a wide combined field-of-regard, while also providing reduction in common volume backscatter. The modularity and desired sensitivity have driven the optical and system design. This paper will detail the modular elements of the FSI design.

NOAA is developing an airborne lidar system for marine fisheries. It measures the direct backscatter of green light from objects, including fish, in the water column. We can, in fact, see fish from an airplane with this lidar. The lidar is described, and examples of fish signatures are presented. Other products are maps of the spatial distribution of fish and vertical profiles of biomass; examples of these are also presented.