We demonstrate a fully submerged underwater LiDAR transceiver system based on single-photon detection technologies. The LiDAR imaging system used a silicon single-photon avalanche diode (SPAD) detector array fabricated in complementary metal-oxide semiconductor (CMOS) technology to measure photon time-of-flight using picosecond resolution time-correlated single-photon counting. The SPAD detector array was directly interfaced to a Graphics Processing Unit (GPU) for real-time image reconstruction capability. Experiments were performed with the transceiver system and target objects immersed in a water tank at a depth of 1.8 meters, with the targets placed at a stand-off distance of approximately 3 meters. The transceiver used a picosecond pulsed laser source with a central wavelength of 532 nm, operating at a repetition rate of 20 MHz and average optical power of up to 52 mW, dependent on scattering conditions. Three-dimensional imaging was demonstrated by implementing a joint surface detection and distance estimation algorithm for real-time processing and visualization, which achieved images of stationary targets with up to 7.5 attenuation lengths between the transceiver and the target. The average processing time per frame was approximately 33 ms, allowing real-time three-dimensional video demonstrations of moving targets at ten frames per second at up to 5.5 attenuation lengths between transceiver and target.

Multivariate extreme value models are used to estimate joint risk in a number of applications, with a particular focus on environmental fields ranging from climatology and hydrology to oceanography and seismic hazards. The semi-parametric conditional extreme value model of Heffernan and Tawn (2004) involving a multivariate regression provides the most suitable of current statistical models in terms of its flexibility to handle a range of extremal dependence classes. However, the standard inference for the joint distribution of the residuals of this model suffers from the curse of dimensionality since in a $d$-dimensional application it involves a $d-1$-dimensional non-parametric density estimator, which requires, for accuracy, a number points and commensurate effort that is exponential in $d$. Furthermore, it does not allow for any partially missing observations to be included and a previous proposal to address this is extremely computationally intensive, making its use prohibitive if the proportion of missing data is non-trivial. We propose to replace the $d-1$-dimensional non-parametric density estimator with a model-based copula with univariate marginal densities estimated using kernel methods. This approach provides statistically and computationally efficient estimates whatever the dimension, $d$ or the degree of missing data. Evidence is presented to show that the benefits of this approach substantially outweigh potential mis-specification errors. The methods are illustrated through the analysis of UK river flow data at a network of 46 sites and assessing the rarity of the 2015 floods in north west England.

This paper discusses the system engineering challenges involved with the transmission of optically encoded data through water. The scenarios of data transmission from an airborne platform to a submerged platform and data transmission from a submerged platform to another submerged platform will be discussed. A photon-counting experimental system was constructed to investigate the transmission of optical data through a 1m long tank of water. This test system incorporated a laser diode operating at a wavelength of 450nm and an optical receiver containing a shallow junction, silicon single photon avalanche diode. The optical data was transmitted through the tank containing similar to 100 litres of water at transmission rates equivalent to 40Mb/s. The attenuation of the optical path was increased by increasing the level of scattering of the photons using Maalox. The effects on the temporal distribution of photons in the optical pulse from adding Maalox are also discussed. The synchronisation of the transmitter and receiver clocks was investigated using reference headers appended to the encoded message signal which the receiver used to correct for timing drift. The performance of this experimental system and experimental results are discussed.

We describe a re-configurable scanning lidar system which can accommodate either a single element detector operating in a scanning mode or a 32 x 32 array detector operating in a non-scanning mode. The system uses a time-of-flight approach in conjunction with the single-photon counting technique to produce 3D images of non-cooperative targets at ranges of greater than one kilometre. Results of data acquired with a single-element detector in a scanning mode at 2.9 km and 4.6 km are reported. The field of view (FoV) was illuminated through a transmitter in a bi-static mode using 125 kHz repetition rate laser pulses at a wavelength of 1550 nm with an average optical power of 0.5W.

This paper reports the performance of a long range 3D imaging system operating at a wavelength of 1550nm incorporating a Geiger mode 32x32 array InGaAs/InP camera. A cross-correlation technique were used to mitigate range aliasing and therefore enable the measurement of the absolute range to single or multiple surfaces within the instantaneous field of view of each pixel in the 2D array. The system uses a fibre amplified laser source operating at an average pulse repetition rate of 125kHz with pulse energies of 2.4 mu J per pulse. Measurements of the absolute range to remote manmade Lambertian surfaces and foliage at ranges up to 10km with range accuracy of better than 4cm are reported. The simultaneous imaging and measurement of the absolute range of two remote manmade Lambertian surfaces separated by >1km is also presented.

We show how nonlinear spectral broadening in a buried chalcogenide mid-infrared waveguide can be used to reshape the spectrum of a femtosecond pulse train at 4260 nm in order to reduce the effects of atmospheric absorption due to carbon dioxide. The nonlinear spectral broadening results in the source with −20 dB spectral width spanning over 3500 nm, from 1700 nm to 5200 nm. This represents a potential route to tailored sources for long-range mid-infrared applications.

We report on progress made in developing a polarimetric and multispectral imaging (MSI) camera that uses a coded filter mask to record spectral data on either a single frame or only a few frames depending on the design of the mask and the application. The design uses commercially available variable filters and beamsplitters to record spectral data simultaneously on a focal plane array, thus negating the need for scanning mechanisms or dispersive optical elements. The data rate depends on the design of the filter mask, pixel resolution, relative motion between target and camera, and the frame rate. The polarimetric data is recorded on a single frame thus eliminating latency. Multispectral data may be recorded on a few frames or a single frame, thus reducing the need for many multi-frame acquisitions. An optimum mask design matched to the imaging lens and focal plane array resolution, allows an object to be detected against spectral clutter or background on a single pixel. Thus a spatially resolved image is not required, alleviating the need for high magnification which would be a practical limitation where compact optical systems are required. The proof-of-principle is reported that uses a 1D variable filter mask to provide spectra of resolved target images in the near field for the visible spectrum. Test data are presented that show the potential for the concept and how it is readily extended to use a 2D mask in a modified design.

Recent developments in time-of-flight, single photon counting, 3D imaging lidar are presented. Preliminary results are presented demonstrating depth profiling and determination of the shape and size of tree canopies at a range of 9km.

The performance of single image deblurring algorithms is typically evaluated via a certain discrepancy measure between the reconstructed image and the ideal sharp image. The choice of metric, however, has been a source of debate and has also led to alternative metrics based on human visual perception. While fixed metrics may fail to capture some small but visible artifacts, perception-based metrics may favor reconstructions with artifacts that are visually pleasant. To overcome these limitations, we propose to assess the quality of reconstructed images via a task-driven metric. In this paper we consider object classification as the task and therefore use the rate of classification as the metric to measure deblurring performance. In our evaluation we use data with different types of blur in two cases: Optical Character Recognition (OCR), where the goal is to recognise characters in a black and white image, and object classification with no restrictions on pose, illumination and orientation. Finally, we show how off-the-shelf classification algorithms benefit from working with deblurred images.

Efficient laser emission in the medium wave infrared (MWIR) is a long established requirement for directed infrared countermeasures (DIRCM). However, until the last decade, there has not been a viable technology for the direct generation of wavelengths in the 3-5μm region and instead indirect methods using optical parametric conversion have been the subject of intense development. Several indirect methods have been developed using different pump wavelengths which represent mature laser technology. For example, 1.54μm from Er:YAG or non-linear conversion from Nd:YAG (1.064μm); 2.1μm from Ho derived from Tm at 1.97μm, or co-doped Ho:Tm in a fibre. These approaches produce the required pump wavelength for efficient 3-5μm generation using either ZGP or OPGaAs [1], however, they are less efficient than direct generation by quantum cascade lasers (QCL). The direct conversion from electrical to optical energy in a QCL is very efficient; wall-plug efficiencies of <10%, depending on wavelength and operating temperature, are typical. High efficiency, together with the high average powers that are now commercially available suggests that the QCL is an attractive laser for DIRCM. However, as protection measures and signal processing techniques advance, one can anticipate that the requirement for sophisticated laser emission in the MWIR becomes more refined. In particular, broadband emission covering a wider, continuous, spectral region will prove harder to counter than that from a few discrete wavelengths. A supercontinuum has been suggested as a possible mechanism for broadband emission. In most investigations into supercontinuum generation, the emphasis has been on producing a wide, flat spectrum covering several hundred nanometres in the visible, near and short wave infrared for stand-off spectroscopic sensing of chemical agents, atmospheric sensing or hyperspectral sensing. These supercontinua are characterised by a spectral bandwidth to pump wavelength ratio of, δλ/λp<1 for a pump wavelength λp in the visible or near infrared. In most applications, the simultaneous generation of a wide spectrum is not required; instead a tuned output suffices. This has the added benefit of improving the efficiency of the laser sensor system since wavelengths which are not required, are not generated. The problem is to understand how a limited continuum might be generated. In the context of DIRCM, the spectral requirement is to produce a controlled spectral emission which matches the 3-5μm atmospheric transmission window. In this paper, a theoretical calculation is presented which shows that a continuous spectrum spanning a few hundred nanometres in the mid infrared (δλ/λp~0.2) can be generated in a simple pump geometry from a mode-locked, ultra-short pulse train using self phase modulation (SPM). Spectral broadening centered on the CO2 absorption band at 4.26μm can be excited to produce all wavelengths for emission in band IV DIRCM. The parameters which affect the spectral output such as pulse power, interaction length, pulse duration and pulse shape are considered for the case where the pump geometry is a collimated beam propagating through a mid infrared glass characterised by a non-linear refractive index n2. The prospects for developing a suitable pump laser are also discussed, in particular, the possibility of using a modelocked QCL.

We report on the performance of a photon-counting optical communication system which was used to transmit optical data at clock rates (not detection rates) of 40Mb/s at a wavelength of 450nm. The transmitted test data patterns comprised of one page of ASCII text preceded by a pseudo-random sequence used as a timing reference pattern by the receiver. The optical data patterns were transmitted through an aquarium tank containing 110 litres of water and were detected at the receiver by a shallow junction silicon single photon avalanche diode detector. An antacid, brand name Maalox, was introduced into the tank to increase the scattering of the optical pulses. The bit error rate and bit rate of the transmitted data were investigated for a range of Maalox concentrations. The optical attenuation and pulse distortion caused by the introduction of Maalox was also investigated.

Time correlated single photon counting (TCSPC) has made tremendous progress during the past ten years enabling improved performance in precision time-of-flight (TOF) rangefinding and lidar. In this review the development and performance of several ranging systems is presented that use TCSPC for accurate ranging and range profiling over distances up to 17km. A range resolution of a few millimetres is routinely achieved over distances of several kilometres. These systems include single wavelength devices operating in the visible; multi-wavelength systems covering the visible and near infra-red; the use of electronic gating to reduce in-band solar background and, most recently, operation at high repetition rates without range aliasing- typically 10MHz over several kilometres. These systems operate at very low optical power (<100μW). The technique therefore has potential for eye-safe lidar monitoring of the environment and obvious military, security and surveillance sensing applications. The review will highlight the theoretical principles of photon counting and progress made in developing absolute ranging techniques that enable high repetition rate data acquisition that avoids range aliasing. Technology trends in TCSPC rangefinding are merging with those of quantum cryptography and its future application to revolutionary quantum imaging provides diverse and exciting research into secure covert sensing, ultra-low power active imaging and quantum rangefinding.

Advances in ultra-short pulse laser technology have resulted in commercially available laser systems capable of generating high peak powers >1GW in tabletop systems. This opens the prospect of generating very wide spectral emissions with a combination of non-linear optical effects in photonic crystal fibres to produce supercontinuua in systems that are readily accessible to military applications. However, military remote sensing rarely requires bandwidths spanning two octaves and it is clear that efficient systems require controlled spectral emission in relevant bands. Furthermore, the limited spectral responsivity of focal plane arrays may impose further restriction on the usable spectrum. A recent innovation which temporally encodes a spectrum using group velocity dispersion allows detection with a photodiode, opening the prospect for high speed hyperspectral sensing and imaging. At the opposite end of the power spectrum, ultra-low power remote sensing using time-correlated single photon counting (SPC) has reduced the laser power requirement and demonstrated remote sensing over 5km during daylight with repetition rates of ~10MHz with ps pulses. Recent research has addressed uncorrelated SPC and waveform transmission to increase data rates for absolute rangefinding whilst avoiding range aliasing. This achievement opens the prospect of combining SPC with high repetition rate temporal encoding of supercontinuua to realise practical hyperspectral remote sensing lidar. The talk will present an overview of these technologies and present a concept which combines them into a single system for high-speed hyperspectral imaging and remote sensing.

The coherence properties of supercontinuum filaments have been studied and several novel techniques for controlling their normally random distribution have been developed

The fundamental characteristics of supercontinuum filament formation in condensed media are investigated along with several novel techniques for controlling their normally random distribution pattern. The utilisation of these supercontinuum filaments as light sources in remote sensing experiments is also discussed. In the first experiments a longitudinal imaging technique is used to scan through the filament and generate a plot of the intensity profile. From this profile an accurate measurement for the filament length and waist may be obtained. Using a similar set-up, the remarkably stable phase relationship is demonstrated between neighbouring filaments and the consequent interference pattern recorded. The second investigations present techniques for controlling both the radial distribution of filaments and the distance at which they form on the axis of propagation.

The effect of beam spatial profile on self-focusing has been investigated. A circular aperture is used to create a Fresnel diffraction pattern. It is shown that self-focusing (a pre-requisite for filament formation) occurs in the presence of the aperture but that no formation is observed when the aperture is removed, even though the beam has higher power well above the threshold for critical power. An analytical solution to the Huygens-Fresnel diffraction integral shows that the axial intensity oscillates between maxima and minima as the distance from the aperture increases and that filament formation coincides with the presence of an axial maximum.

This paper presents accurate measurements of the filament length and width by various experimental techniques. Coherence property is also demonstrated by using an array of diffractive microlenses to focus the beam through a sample of B 270 glass. Finally, a simple technique is presented for inducing filamentation without the aid of a focusing lens and for controlling the location of the filament on the axis of propagation. This study is essential to gain an understanding of the various physical properties which affect the filamentation process.