Ling-An Wu

Ultraviolet (UV) imaging can reveal structural flaws, fluorescent markers, and characteristic signals generated by material changes that are difficult to detect in visible-light imaging and has been applied in numerous fields. However, in traditional UV imaging, it is challenging to simultaneously acquire reflection/transmission and UV-excited fluorescence imaging, often requiring additional optical filters to eliminate interference bands. Here, we demonstrate a dual-mode UV lensless imaging method based on diffuser speckle modulation, which enables direct reconstruction of both reflection/transmission and fluorescence images from a single-shot measurement. Compared with conventional methods, our scheme eliminates the need for optical lenses and filters, enabling an ultra-compact lensless system. This approach can be further integrated with other image post-processing operations or general computer vision tasks for the development of task-oriented intelligent UV cameras.

Although “ghost” imaging has seen remarkable advances since it was first studied 30 years ago, there have been few studies on x-ray ghost imaging (XGI) due to technical difficulties. A chief potential advantage of XGI is that imaging can be achieved with much lower radiation intensity. However, in previous XGI experiments with continuous wave x-ray illumination, the total sampling time was actually quite long, leading to additional radiation dose accumulation. In this paper, we report a proof-of-concept experiment using a high-sensitivity single-pixel detector together with a simple tabletop x-ray source followed by an adjustable mechanical shutter, which could reduce the total exposure time to as little as 5 ms. Imaging was realized with a spatial resolution of 100 μm under a low radiation dose. The influence of x-ray tube current, sampling frequency, and exposure times on image quality was examined. Our experiment illustrates a practical method for low-radiation dose ghost imaging and provides valuable reference data for the future development of medical XGI cameras.

For practical applications of X-ray ghost imaging (XGI), the imaging time is a major challenge. In this paper, we propose a fast XGI scheme based on a continuous translation mask with etched aggregate patterns. High contrast-to-noise ratio images are obtained with an exposure time of only 4 s and 24 s for a 3.4 mm×3.8 mm and 5.9 mm×6.1 mm field-of-view, respectively. The spatial resolution can reach ∼150 μm. The influences of the sampling frequency, the mask scanning speed, and the detector integration time on image quality are examined, from which we demonstrate that the imaging time can be further reduced by increasing the mask translation speed. By applying a compressed sensing reconstruction algorithm, the exposure time is greatly reduced while maintaining image quality. Our method indicates a path for the development of future XGI applications.

Single-pixel imaging (SPI) is a technique that employs a non-scanning single-pixel detector to reconstruct the image. Research has mostly concentrated on static targets, as image reconstruction requires multiple samples, and SPI of moving objects, especially multiple moving objects, remains challenging. Here we propose a motion compensation scheme for SPI of multiple moving targets without prior motion information, allowing image reconstruction from any selected frame of the motion process. By obtaining the number and spatial positions of moving targets, the corresponding regions of the imaging modulation patterns are shifted to the target positions of any selected frame to reconstruct the images, thereby achieving high frame rate multi-target motion imaging. In laboratory experiments, with a digital micromirror device (DMD) modulated at a frequency of 20 kHz, the imaging frame rate reached 161 Hz for a frame size of 128×128 pixels. The peak signal-to-noise ratio of each frame improved by at least 3.87 dB and up to 6.54 dB after motion compensation. In addition, outdoor field experiments using an SPI prototype have successfully realized long-distance (258.5 m) multi-target motion imaging; with a DMD modulated at a frequency of 8 kHz, the imaging frame rate reached 48 Hz for a frame size of 128×128 pixels, and a contrast-to-noise ratio improvement of over 10 dB was achieved after compensation.

Single-pixel imaging (SPI), especially when integrated with deep neural networks like deep image prior networks (DIP-Net) or data-driven networks (DD-Net), has gained considerable attention for its capability to generate high-quality reconstructed images, even in the presence of sub-sampling conditions. However, DIP-Net often requires thousands of iterations to achieve high-quality image reconstruction, and DD-Net performs optimally only when the target closely resembles the features present in its training set. To overcome these limitations, we propose a dual-network iterative optimization (SPI-DNIO) framework that combines the strengths of both DD-Net and DIP-Net. It has been demonstrated that this approach can recover high-quality images with fewer iteration steps. Furthermore, to address the challenge of SPI inputs having less effective information at low sampling rates, we have designed a residual block enriched with gradient information, which can convey details to deeper layers, thereby enhancing the deep networks learning capabilities. We have applied these techniques to both indoor experiments with active lighting and outdoor long-range experiments with passive lighting. Our experimental results confirm the exceptional reconstruction capabilities and generalization performance of the SPI-DNIO framework.

We present a proof-of-principle demonstration of energy-resolved resonant neutron ghost imaging. Based on the resonant absorption dips of different elements, we simultaneously image and distinguish the composition of three differently shaped components of an object. The initial neutron beam is spatially and energy selectively modulated by a series of Hadamard matrix masks of pixel width 100 m. The spectral intensity transmitted through an object is measured by a ⁶Li glass single-pixel detector. Through integration of the total counts within resonant dips and correlating them with the corresponding Hadamard patterns, isotope-specific images of In, Ag and W objects are obtained at an effective spatial resolution of ~200 m. Reconstruction algorithms based on compressed sensing or convolutional neural networks can greatly reduce the data acquisition time by ~70% with respect to the full set of 1024 patterns, as well as enhance the image quality. Incorporating ghost imaging into energy-resolved neutron imaging thus has great potential for the simultaneous realization of fine spatial and spectral resolution, which has important value for the noninvasive analysis of material composition and distribution not only in basic research but also in industrial applications.

Spinning coding masks, recognized for their fast modulation rate and cost-effectiveness, are now often used in real-time single-pixel imaging (SPI). However, in the photon-counting regime, they encounter difficulties in synchronization between the coding mask patterns and the photon detector, unlike digital micromirror devices. To address this issue, we propose a scheme that assumes a constant disk rotation speed throughout each cycle and models photon detection as a non-homogeneous Poisson process (NHPP). This effectively resolves synchronization problems and compensates for speed fluctuations. To validate this method, we designed and fabricated a single-pixel camera prototype that can capture images under an illumination of less than one photon per pixel, with a modulation rate of approximately 100 kHz and an imaging speed of 28 frames per second. The camera is compact, lightweight, and low cost and should find many practical applications for imaging under extremely low-light conditions.

Single-pixel imaging (SPI) has gained considerable attention over the past two decades but is still inadequate for imaging moving objects in practical scenarios. Recently, spinning masks have been employed for real-time SPI due to their fast modulation speeds. However, the intensity fluctuations caused by ambient light can overlap with the desired modulation from the mask, leading to a significant degradation in image quality, making it unable to function properly in environments with unstable lighting. Here we propose a rotating disc encoded with cyclic S-matrices, leveraging both reflective and transmissive signals for differential detection. Simulations and experiments demonstrate interference-resistant real-time SPI, achieving 31 fps at about 60 kHz modulation rates. This approach effectively suppresses external noise and is extendable to infrared and terahertz bands, advancing SPI’s practical applications.