Computational version of the correlation light-field camera.

Light-field cameras allow the acquisition of both the spatial and angular components of the light-field. The conventional way to perform such acquisitions leads to a strong spatio-angular resolution limitation but correlation-enabled plenoptic cameras have been introduced recently that relax this constraint. Here we use a computational version of this concept to acquire realistic light-fields images using a commercial DSLR Camera lens as an imaging system. By placing the image sensor in the focal plane of a lens, within the camera we ensure the acquisition of pure angular components together with the spatial information. We perform an acquisition presenting a high spatio-angular rays resolution obtained through a trade off of the temporal resolution. The acquisition reported is photo-realistic and the acquisition of diffraction limited features is observed with the setup. Finally, we demonstrate the refocusing abilities of the camera.

Dual-band single-pixel telescope.

Single-pixel imaging systems can obtain images from a wide range of wavelengths at low-cost compared to those using conventional multi-pixel, focal-plane array sensors, especially at wavelengths outside the visible spectrum. The ability to sense short-wave infrared radiation with single-pixel techniques extends imaging capability to adverse weather conditions and environments, such as fog, haze, or night time. In this work, we demonstrate a dual-band single-pixel telescope for imaging at both visible (VIS) and short-wave infrared (SWIR) spectral regions simultaneously under some of these outdoor weather conditions. At 64 × 64 pixel-resolution, our system has achieved continuous VIS and SWIR imaging of various objects at a frame rate up to 2.4 Hz. Visual and contrast comparison between the reconstructed VIS and SWIR images emphasizes the significant contribution of infrared observation using the single-pixel technique. The single-pixel telescope provides an alternative cost-effective imaging solution for synchronized dual-waveband optical applications.

A compact acoustic spanner to rotate macroscopic objects.

Waves can carry both linear and angular momentum. When the wave is transverse (e.g. light), the angular momentum can be characterised by the "spin" angular momentum associated with circular polarisation, and the "orbital" angular momentum (OAM) arising from the phase cross-section of the beam. When the wave is longitudinal (e.g. sound) there is no polarization and hence no spin angular momentum. However, a suitably phase-structured sound beam can still carry OAM. Observing the transfer of OAM from sound to a macroscopic object provides an excellent opportunity to study the exchange of energy between waves and matter. In this paper we show how to build a compact free-space acoustic spanner based on a 3D-printed sound-guiding structure and common electronic components. We first characterise the sound fields by measuring both phase and amplitude maps, and then show a video of our free-space acoustic spanner in action, in which macroscopic objects spin in a circular motion and change direction of rotation according to the handedness of the OAM acoustic field.

Hybrid 3D ranging and velocity tracking system combining multi-view cameras and simple LiDAR.

Scanning our surroundings has become one of the key challenges in automation. Effective and efficient position, distance and velocity sensing is key to accurate decision making in automated applications from robotics to driverless cars. Light detection and ranging (LiDAR) has become a key tool in these 3D sensing applications, where the time-of-flight (TOF) of photons is used to recover distance information. These systems typically rely on scanning of a laser spot to recover position information. Here we demonstrate a hybrid LiDAR approach which combines a multi-view camera system for position and distance information, and a simple (scanless) LiDAR system for velocity tracking and depth accuracy. We show that we are able to combine data from the two component systems to provide a compound image of a scene with position, depth and velocity data at more than 1 frame per second with depth accuracy of 2.5 cm or better. This hybrid approach avoids the bulk and expense of scanning systems while adding velocity information. We hope that this approach will offer a simpler, more robust alternative to 3D scanning systems for autonomous vehicles.

Deep learning for real-time single-pixel video.

Single-pixel cameras capture images without the requirement for a multi-pixel sensor, enabling the use of state-of-the-art detector technologies and providing a potentially low-cost solution for sensing beyond the visible spectrum. One limitation of single-pixel cameras is the inherent trade-off between image resolution and frame rate, with current compressive (compressed) sensing techniques being unable to support real-time video. In this work we demonstrate the application of deep learning with convolutional auto-encoder networks to recover real-time 128 × 128 pixel video at 30 frames-per-second from a single-pixel camera sampling at a compression ratio of 2%. In addition, by training the network on a large database of images we are able to optimise the first layer of the convolutional network, equivalent to optimising the basis used for scanning the image intensities. This work develops and implements a novel approach to solving the inverse problem for single-pixel cameras efficiently and represents a significant step towards real-time operation of computational imagers. By learning from examples in a particular context, our approach opens up the possibility of high resolution for task-specific adaptation, with importance for applications in gas sensing, 3D imaging and metrology.

Sub-shot-noise shadow sensing with quantum correlations.

The quantised nature of the electromagnetic field sets the classical limit to the sensitivity of position measurements. However, techniques based on the properties of quantum states can be exploited to accurately measure the relative displacement of a physical object beyond this classical limit. In this work, we use a simple scheme based on the split-detection of quantum correlations to measure the position of a shadow at the single-photon light level, with a precision that exceeds the shot-noise limit. This result is obtained by analysing the correlated signals of bi-photon pairs, created in parametric downconversion and detected by an electron multiplying CCD (EMCCD) camera employed as a split-detector. By comparing the measured statistics of spatially anticorrelated and uncorrelated photons we were able to observe a significant noise reduction corresponding to an improvement in position sensitivity of up to 17% (0.8dB). Our straightforward approach to sub-shot-noise position measurement is compatible with conventional shadow-sensing techniques based on the split-detection of light-fields, and yields an improvement that scales favourably with the detector's quantum efficiency.

A Russian Dolls ordering of the Hadamard basis for compressive single-pixel imaging.

Single-pixel imaging is an alternate imaging technique particularly well-suited to imaging modalities such as hyper-spectral imaging, depth mapping, 3D profiling. However, the single-pixel technique requires sequential measurements resulting in a trade-off between spatial resolution and acquisition time, limiting real-time video applications to relatively low resolutions. Compressed sensing techniques can be used to improve this trade-off. However, in this low resolution regime, conventional compressed sensing techniques have limited impact due to lack of sparsity in the datasets. Here we present an alternative compressed sensing method in which we optimize the measurement order of the Hadamard basis, such that at discretized increments we obtain complete sampling for different spatial resolutions. In addition, this method uses deterministic acquisition, rather than the randomized sampling used in conventional compressed sensing. This so-called 'Russian Dolls' ordering also benefits from minimal computational overhead for image reconstruction. We find that this compressive approach performs as well as other compressive sensing techniques with greatly simplified post processing, resulting in significantly faster image reconstruction. Therefore, the proposed method may be useful for single-pixel imaging in the low resolution, high-frame rate regime, or video-rate acquisition.

Adaptive foveated single-pixel imaging with dynamic supersampling.

In contrast to conventional multipixel cameras, single-pixel cameras capture images using a single detector that measures the correlations between the scene and a set of patterns. However, these systems typically exhibit low frame rates, because to fully sample a scene in this way requires at least the same number of correlation measurements as the number of pixels in the reconstructed image. To mitigate this, a range of compressive sensing techniques have been developed which use a priori knowledge to reconstruct images from an undersampled measurement set. Here, we take a different approach and adopt a strategy inspired by the foveated vision found in the animal kingdom-a framework that exploits the spatiotemporal redundancy of many dynamic scenes. In our system, a high-resolution foveal region tracks motion within the scene, yet unlike a simple zoom, every frame delivers new spatial information from across the entire field of view. This strategy rapidly records the detail of quickly changing features in the scene while simultaneously accumulating detail of more slowly evolving regions over several consecutive frames. This architecture provides video streams in which both the resolution and exposure time spatially vary and adapt dynamically in response to the evolution of the scene. The degree of local frame rate enhancement is scene-dependent, but here, we demonstrate a factor of 4, thereby helping to mitigate one of the main drawbacks of single-pixel imaging techniques. The methods described here complement existing compressive sensing approaches and may be applied to enhance computational imagers that rely on sequential correlation measurements.

Real-time imaging of methane gas leaks using a single-pixel camera.

We demonstrate a camera which can image methane gas at video rates, using only a single-pixel detector and structured illumination. The light source is an infrared laser diode operating at 1.651µm tuned to an absorption line of methane gas. The light is structured using an addressable micromirror array to pattern the laser output with a sequence of Hadamard masks. The resulting backscattered light is recorded using a single-pixel InGaAs detector which provides a measure of the correlation between the projected patterns and the gas distribution in the scene. Knowledge of this correlation and the patterns allows an image to be reconstructed of the gas in the scene. For the application of locating gas leaks the frame rate of the camera is of primary importance, which in this case is inversely proportional to the square of the linear resolution. Here we demonstrate gas imaging at ~25 fps while using 256 mask patterns (corresponding to an image resolution of 16×16). To aid the task of locating the source of the gas emission, we overlay an upsampled and smoothed image of the low-resolution gas image onto a high-resolution color image of the scene, recorded using a standard CMOS camera. We demonstrate for an illumination of only 5mW across the field-of-view imaging of a methane gas leak of ~0.2 litres/minute from a distance of ~1 metre.

Improving the signal-to-noise ratio of single-pixel imaging using digital microscanning.

Single-pixel cameras provide a means to perform imaging at wavelengths where pixelated detector arrays are expensive or limited. The image is reconstructed from measurements of the correlation between the scene and a series of masks. Although there has been much research in the field in recent years, the fact that the signal-to-noise ratio (SNR) scales poorly with increasing resolution has been one of the main limitations prohibiting the uptake of such systems. Microscanning is a technique that provides a final higher resolution image by combining multiple images of a lower resolution. Each of these low resolution images is subject to a sub-pixel sized lateral displacement. In this work we apply a digital microscanning approach to an infrared single-pixel camera. Our approach requires no additional hardware, but is achieved simply by using a modified set of masks. Compared to the conventional Hadamard based single-pixel imaging scheme, our proposed framework improves the SNR of reconstructed images by ∼ 50 % for the same acquisition time. In addition, this strategy also provides access to a stream of low-resolution 'preview' images throughout each high-resolution acquisition.

Single-pixel three-dimensional imaging with time-based depth resolution.

Time-of-flight three-dimensional imaging is an important tool for applications such as object recognition and remote sensing. Conventional time-of-flight three-dimensional imaging systems frequently use a raster scanned laser to measure the range of each pixel in the scene sequentially. Here we show a modified time-of-flight three-dimensional imaging system, which can use compressed sensing techniques to reduce acquisition times, whilst distributing the optical illumination over the full field of view. Our system is based on a single-pixel camera using short-pulsed structured illumination and a high-speed photodiode, and is capable of reconstructing 128 × 128-pixel resolution three-dimensional scenes to an accuracy of ∼3 mm at a range of ∼5 m. Furthermore, by using a compressive sampling strategy, we demonstrate continuous real-time three-dimensional video with a frame-rate up to 12 Hz. The simplicity of the system hardware could enable low-cost three-dimensional imaging devices for precision ranging at wavelengths beyond the visible spectrum.

A fast 3D reconstruction system with a low-cost camera accessory.

Photometric stereo is a three dimensional (3D) imaging technique that uses multiple 2D images, obtained from a fixed camera perspective, with different illumination directions. Compared to other 3D imaging methods such as geometry modeling and 3D-scanning, it comes with a number of advantages, such as having a simple and efficient reconstruction routine. In this work, we describe a low-cost accessory to a commercial digital single-lens reflex (DSLR) camera system allowing fast reconstruction of 3D objects using photometric stereo. The accessory consists of four white LED lights fixed to the lens of a commercial DSLR camera and a USB programmable controller board to sequentially control the illumination. 3D images are derived for different objects with varying geometric complexity and results are presented, showing a typical height error of <3 mm for a 50 mm sized object.

Simultaneous real-time visible and infrared video with single-pixel detectors.

Conventional cameras rely upon a pixelated sensor to provide spatial resolution. An alternative approach replaces the sensor with a pixelated transmission mask encoded with a series of binary patterns. Combining knowledge of the series of patterns and the associated filtered intensities, measured by single-pixel detectors, allows an image to be deduced through data inversion. In this work we extend the concept of a 'single-pixel camera' to provide continuous real-time video at 10 Hz , simultaneously in the visible and short-wave infrared, using an efficient computer algorithm. We demonstrate our camera for imaging through smoke, through a tinted screen, whilst performing compressive sampling and recovering high-resolution detail by arbitrarily controlling the pixel-binning of the masks. We anticipate real-time single-pixel video cameras to have considerable importance where pixelated sensors are limited, allowing for low-cost, non-visible imaging systems in applications such as night-vision, gas sensing and medical diagnostics.

Fast full-color computational imaging with single-pixel detectors.

Single-pixel detectors can be used as imaging devices by making use of structured illumination. These systems work by correlating a changing incident light field with signals measured on a photodiode to derive an image of an object. In this work we demonstrate a system that utilizes a digital light projector to illuminate a scene with approximately 1300 different light patterns every second and correlate these with the back scattered light measured by three spectrally-filtered single-pixel photodetectors to produce a full-color high-quality image in a few seconds of data acquisition. We utilize a differential light projection method to self normalize the measured signals, improving the reconstruction quality whilst making the system robust to external sources of noise. This technique can readily be extended for imaging applications at non-visible wavebands.

Waveguide grating mirror in a fully suspended 10 meter Fabry-Perot cavity.

We report on the first demonstration of a fully suspended 10 m Fabry-Perot cavity incorporating a waveguide grating as the coupling mirror. The cavity was kept on resonance by reading out the length fluctuations via the Pound-Drever-Hall method and employing feedback to the laser frequency. From the achieved finesse of 790 the grating reflectivity was determined to exceed 99.2% at the laser wavelength of 1064 nm, which is in good agreement with rigorous simulations. Our waveguide grating design was based on tantala and fused silica and included a ≈ 20 nm thin etch stop layer made of Al2O3 that allowed us to define the grating depth accurately and preserve the waveguide thickness during the fabrication process. Demonstrating stable operation of a waveguide grating featuring high reflectivity in a suspended low-noise cavity, our work paves the way for the potential application of waveguide gratings as mirrors in high-precision interferometry, for instance in future gravitational wave observatories.