Single-pixel Imaging Technique Research Articles

Single-pixel imaging through non-homogeneous turbid media with adaptive illumination

The presence of scattering media limits the quality of images obtained by optical systems. Single-pixel imaging techniques based on structured illumination are highly tolerant to the presence of scattering between the object and the sensor, but very sensitive when the scattering medium is between the light source and the object. This makes it difficult to develop single-pixel imaging techniques for the case of objects immersed in scattering media. We present what we believe to be a new system for imaging objects through inhomogeneous scattering media in an epi-illumination configuration. It works in an adaptive way by combining diffuse optical imaging (DOI) and single pixel imaging (SPI) techniques in two stages. First, the turbid media is characterized by projecting light patterns with an LED array and applying DOI techniques. Second, the LED array is programmed to project light only through the less scattering areas of the media, while simultaneously using a digital micromirror device (DMD) to project light patterns onto the target using Hadamard basis coding functions. With this adaptive technique, we are able to obtain images of targets through two different scattering media with better quality than using conventional illumination. We also show that the system works with fluorescent targets.

A primal-dual data-driven method for computational optical imaging with a photonic lantern.

Optical fibers aim to image in vivo biological processes. In this context, high spatial resolution and stability to fiber movements are key to enable decision-making processes (e.g. for microendoscopy). Recently, a single-pixel imaging technique based on a multicore fiber photonic lantern has been designed, named computational optical imaging using a lantern (COIL). A proximal algorithm based on a sparsity prior, dubbed SARA-COIL, has been further proposed to solve the associated inverse problem, to enable image reconstructions for high resolution COIL microendoscopy. In this work, we develop a data-driven approach for COIL. We replace the sparsity prior in the proximal algorithm by a learned denoiser, leading to a plug-and-play (PnP) algorithm. The resulting PnP method, based on a proximal primal-dual algorithm, enables to solve the Morozov formulation of the inverse problem. We use recent results in learning theory to train a network with desirable Lipschitz properties, and we show that the resulting primal-dual PnP algorithm converges to a solution to a monotone inclusion problem. Our simulations highlight that the proposed data-driven approach improves the reconstruction quality over variational SARA-COIL method on both simulated and real data.

Vector-guided Fourier single-pixel imaging.

The Fourier single-pixel imaging technique exhibits great potential for compressive imaging. However, the utilization of low sampling ratio can introduce unwanted ringing artifacts, thereby compromising the fidelity of reconstructed image detail. To address this issue, Vector guided Fourier single-pixel imaging (V-FSI) has been proposed. We analyze the statistical properties in the edge vector field derived from images with low sampling ratio. Based on this information, a tailored sampling map is designed to acquire the significant high-frequency components for image reconstruction. Experimental results demonstrate the remarkable effectiveness of the proposed V-FSI method in enhancing image quality. Notably, V-FSI exhibits exceptional capabilities in perceiving and preserving the details of the objects, particularly for objects characterized by pronounced periodicity and directionality.

Physics-informed and simulation-driven optimization for binary Fourier single-pixel imaging.

Binary patterns are used in fast Fourier single-pixel imaging (FSI) technology to increase the imaging speed at the expense of spatial resolution or image quality. In this Letter, we propose a method for optimizing the image quality-speed trade-off that is informed by physical principles and driven by data from simulations. To compensate for the quantization error induced by binary dithering, convolution kernels are proposed and optimized for both low and high spatial frequencies. The proposed method has been demonstrated to work in both simulation and experiments. Other single-pixel imaging (SPI) techniques may also benefit from this approach.

Efficient and noise-resistant single-pixel imaging based on Pseudo-Zernike moments.

An efficient and noise-resistant single-pixel imaging (SPI) technique based on Pseudo-Zernike moments (PZ-SPI) is proposed. In this technique, the illumination light fields are modulated to satisfy the Pseudo-Zernike polynomials. Then the modulated light fields are projected onto the object. And the single-pixel detector is used to measure the reflected light intensities to calculate the Pseudo-Zernike moments. Finally, the object image is reconstructed by iterative summation of the product of the Pseudo-Zernike polynomials and the Pseudo-Zernike moments. Through the numerical simulation and experimental demonstration, PZ-SPI can effectively reconstruct image at low sampling ratios. Besides, comparing with the Fourier-SPI and Zernike-SPI, PZ-SPI has good robustness to background noise in SPI system. These advantages expand the application of PZ-SPI in complex environments.

Extended Kalman filtering for enhancing a cyclic pattern-updating scheme of single-pixel dynamic fluorescence spatial frequency domain imaging.

A single-pixel imaging technique applied to fluorescence spatial frequency domain imaging (f-SFDI) brings many positive benefits, but its low frame rate will also lead to severe quantitative degradation when dynamically imaging a vibrant target. This work presents a novel, to the best of our knowledge, single-pixel imaging approach that combines the extended Kalman filtering (EKF) and a cyclic one-pattern updating for an enhanced dynamic f-SFDI. The cyclic one-pattern updating scheme enables the dynamic imaging at a high frame rate, and on this basis, the imaging process of an intensity temporally varying target (assuming no structure motion in the scene) is dynamically modelled, and accordingly, the surface intensities and images at each sampling time point simultaneously estimated via the EKF. Simulation and phantom validations demonstrate that the method can improve the quantitative accuracy of the results. An in vivo experiment performed on two mice for dynamic monitoring of photosensitizer doses in a photodynamic therapy further demonstrates the clinical feasibility of the proposed method.

Image denoising and deringing for fourier single-pixel imaging based on upgraded weighted nuclear norm minimization

Single-pixel imaging techniques have been extensively studied because of the low requirements for the detector resolution and the low equipment cost. However, a large amount of noise and ringing commonly occur in the reconstructed images via previous single-pixel imaging methods at ultra-low sampling rates, which limits the real applications. In this paper, a novel Fourier single-pixel imaging method is proposed to reconstruct high-quality single-pixel imaging images at ultra-low sampling rates, which involves the stages of Fourier single-pixel imaging, block matching, upgraded weight nuclear norm minimization, and aggregation. To decrease the computational burden while important image details, block matching and aggregation are employed for the blocks divided from the low-quality image reconstructed by Fourier single-pixel imaging. To adapt the proposed upgraded weight nuclear norm minimization to different contaminated images, an adaptive scheme for the iterative regularization parameter is proposed to update the traditional weight nuclear norm minimization. Experimental results indicate that the proposed Fourier single-pixel imaging method based on upgraded weight nuclear norm minimization exhibits the capacity to reliably and effectively reconstruct high-quality images for single-pixel imaging at low sampling rates, even at the sampling rate of 1 %. Experimental results also demonstrate that the proposed method performs better image reconstruction than some existing single-pixel imaging.

Single-pixel detecting of rotating object using Zernike illumination

In this paper, a single-pixel imaging technique based on Zernike illumination is firstly introduced to detect rotating object. By measuring the specific Zernike moments through a single-pixel detector, the rotation angle of rotating object can be obtained. This method requires only four differential Zernike patterns to illuminate a rotating object in one frame. And the Zernike moments can be calculated from the corresponding reflected intensities collected by the single-pixel detector. Then the rotation angle can be calculated from the phase change of Zernike moments due to the rotation invariant. By theoretical simulation and experimental demonstration, the Zernike moments of the targets are successfully obtained through single-pixel imaging system. The rotation angles are calculated from the measured Zernike moments and agree well with the actual rotation angles. The errors of simulated rotation angle are below 1°. And the RMSEs of the experimental results are below 3° for binary images and below 4° for gray images. The proposed scheme provides a new method for applications on detecting rotating object.

Optimal Reconstruction of Single-Pixel Images through Feature Feedback Mechanism and Attention

The single-pixel imaging technique can reconstruct high-quality images using only a bucket detector with no spatial resolution, and the image quality is degraded in order to meet the demands of real-time applications. According to some studies of algorithm performance, the network model performs differently in simulated and real-world experiments. We propose an end-to-end neural network capable of reconstructing 2D images from experimentally obtained 1D signals optimally. In order to improve the image quality of real-time single-pixel imaging, we built a feedback module in the hidden layer of the recurrent neural network to implement feature feedback. The feedback module fuses high-level features of undersampled images with low-level features through dense jump connections and multi-scale balanced attention modules to gradually optimize the feature extraction process and reconstruct high-quality images. In addition, we introduce a learning strategy that combines mean loss with frequency domain loss to improve the network’s ability to reconstruct complex undersampled images. In this paper, the factors that lead to the degradation of single-pixel imaging are analyzed, and a network degradation model suitable for physical imaging systems is designed. The experiment results indicate that the reconstructed images utilizing the proposed method have better quality metrics and visual effects than the excellent methods in the field of single-pixel imaging.

Rotationally synchronized single-pixel imaging for a fast-rotating object

In practical application environments, objects are rarely stationary, which makes it difficult to image dynamic objects with conventional single-pixel imaging (SI) techniques. In this paper, a rotationally synchronized single-pixel imaging (RS-SI) method is proposed to image a dynamic object in rotation. The modulation pattern rotates in sync with the rapidly rotating object, and the center of object rotation is ensured to be in line with the center of the illumination pattern. Then, RS-SI reconstruction algorithm is used to reconstruct the image of the rotating object by collecting the light reflected from the object's surface. This method does not require advanced knowledge of the object rotation speed to complete the imaging. Simulation and experimental results confirm that the RS-SI can reconstruct images of an object at rotational speeds up to 422.0 rpm and can also complete imaging of a variable-speed rotating object. Imaging results of 128 × 128 pixels at a sampling rate of 27.47% can be obtained with high fidelity. The proposed RS-SI is a method for imaging an object in rotation, providing insights for future applications of single-pixel imaging technology for defect detection in rotating parts.

Masked autoencoder for highly compressed single-pixel imaging.

The single-pixel imaging technique uses multiple patterns to modulate the entire scene and then reconstructs a two-dimensional (2-D) image from the single-pixel measurements. Inspired by the statistical redundancy of natural images that distinct regions of an image contain similar information, we report a highly compressed single-pixel imaging technique with a decreased sampling ratio. This technique superimposes an occluded mask onto modulation patterns, realizing that only the unmasked region of the scene is modulated and acquired. In this way, we can effectively decrease 75% modulation patterns experimentally. To reconstruct the entire image, we designed a highly sparse input and extrapolation network consisting of two modules: the first module reconstructs the unmasked region from one-dimensional (1-D) measurements, and the second module recovers the entire scene image by extrapolation from the neighboring unmasked region. Simulation and experimental results validate that sampling 25% of the region is enough to reconstruct the whole scene. Our technique exhibits significant improvements in peak signal-to-noise ratio (PSNR) of 1.5 dB and structural similarity index measure (SSIM) of 0.2 when compared with conventional methods at the same sampling ratios. The proposed technique can be widely applied in various resource-limited platforms and occluded scene imaging.

Feature single-pixel imaging: What you see is what you want

Traditional imaging technology is basically a “what you see is what you get” imaging method, providing images that resemble human vision for post-information acquisition and interpretation. This Letter proposes a feature single-pixel imaging technique that allows for direct imaging of specific features without the need for traditional image processing methods. This approach enables the imaging of a single feature at a remarkable ultra-low sampling ratio of 2% with a resolution of up to 128 × 128 pixels, operating at a frequency of 67 Hz, even in dynamic environments where the location of the target or background features may change. Additionally, the proposed method demonstrates the ability to selectively image defects on an integrated circuit wafer. This research offers a significant advancement in the development of single-pixel imaging for feature information acquisition and has immense potential for applications in various industries and daily life scenarios.

DCT single-pixel detecting for wavefront measurement

In this work, a checkerboard pattern is used to divide an unknown field into the signal and reference parts, and based on the interference between them, we propose and demonstrate a discrete cosine transform single-pixel imaging (DCT-SPI) technique for measuring the spatial wavefront of an unknown field. Here, the discrete cosine transform (DCT) basis was used for generating the modulation patterns to acquire the spectrum of the wavefront by the phase-shifting technique, and the wavefront to be measured can be retrieved by applying a second-order correlation calculation. We first demonstrate the proposed DCT-SPI can be used for measuring the spatial wavefronts simulated by an amplitude SLM(A-SLM). As compared to the phase-shifting Fourier single-pixel imaging (FSPI) technique, the imaging speed of DCT-SPI is four times faster. In addition, under the RAM capacity limit of a digital micromirror device (DMD), DCT-SPI can realize a higher resolution. Further, we experimentally demonstrate that DCT-SPI can also be used for measuring the phase and amplitude of a dragonfly wing, which suggests it can be applied in the field of quantitative phase imaging with a higher speed and imaging resolution as compared with FSPI.

High-quality reconstruction of single-pixel imaging using discrete W transform

Single-pixel imaging (SPI) achieves numerous advantages over traditional imaging methods, such as multi-wavelength detection and non-local imaging. In this paper, we report a novel single-pixel imaging technique based on discrete W transform (DWT). In this method, we adopt the star sampling strategy to generate two sets of W basis patterns, these patterns are used to differentially modulate the light source, then the spectrum of the target object is acquired from the light intensities measured by a single-pixel detector. Finally, inverse DWT is utilized the to reconstruct the image. Therefore, high-quality images can be retrieved quickly from sub-Nyquist measurements compared with other transform domain SPI methods. Simulation and experimental results validate the effectiveness of the proposed technique.

A Single-Pixel High-Precision Imaging Technique Based on a Discrete Zernike Transform for High-Efficiency Image Reconstructions

Single-pixel imaging (SPI) has attracted increasing attention in recent years because of its advantages in imaging systems. However, a low reconstruction quality and a long reconstruction time have hindered the development of SPI. Hence, in this study, we propose a Zernike SPI (ZSPI) technique to reduce the number of illumination patterns and reconstruction time whilst retaining robustness. First, the ZSPI technique was theoretically demonstrated. Phase-shifting Zernike moment projections were used to illuminate the target and an inverse Zernike transform was used to reconstruct the desired image. In order to prove the feasibility, numerical simulations were carried out with different sample ratios (SRs) ranging from 0.1 to 0.3; an acceptable reconstruction appeared at approximately 0.1. This result indicated that ZSPI could obtain satisfactory reconstruction results at low SRs. Further simulation and physical experiments were compared with different reconstruction algorithms, including noniterative, linear iterative, and nonlinear iterative methods under speckle modulation patterns at a sample of 0.1 in terms of different targets. The results revealed that ZSPI had a higher imaging quality and required less imaging time, particularly for low-frequency targets. The method presented in this study has advantages for the high-efficiency imaging of low-frequency targets, which can provide a new solution for the SPI method.

VGenNet: Variable Generative Prior Enhanced Single Pixel Imaging

Single-pixel imaging (SPI) is an emerging imaging methodology that converts a two- or even three-dimensional image acquisition problem into a one-dimensional (1D) temporal-signal detection problem. Thus, it is crucially important to develop efficient SPI techniques for image reconstruction from the 1D measurements, in particular, an undersampled one. Recently, various studies have demonstrated the superiority of deep learning for SPI. However, due to the generalization issue, conventional data-driven deep learning is a task-specific approach. One needs to retrain the neural network for different SPI imaging problems and different types of objects. Here, we propose a variable generative network enhanced SPI algorithm (VGenNet) by incorporating a model-driven fine-tuning process into a generative model that may have been trained for other tasks. VGenNet simultaneously updates the input vector and the weights in a generator to generate feasible solutions that reproduce the raw measurements. We demonstrate the proposed technique with indoor SPI and outdoor 3D single-pixel LiDAR experiments. Our results show that high-quality images can be reconstructed at low sampling ratios under different system configurations, demonstrating the good performance and flexibility of VGenNet. Overall, the proposed VGenNet is a general framework to take advantage of both the data and physics priors, allowing the direct use of a pretrained generative model to solve various inverse imaging problems.

Interaction-free, single-pixel quantum imaging with undetected photons

A typical imaging scenario requires three basic ingredients: (1) a light source that emits light, which in turn interacts and scatters off the object of interest; (2) detection of the light being scattered from the object and (3) a detector with spatial resolution. These indispensable ingredients in typical imaging scenarios may limit their applicability in the imaging of biological or other sensitive specimens due to unavailable photon-starved detection capabilities and inevitable damage induced by interaction. Here, we propose and experimentally realize a quantum imaging protocol that alleviates all three requirements. By embedding a single-photon Michelson interferometer into a nonlinear interferometer based on induced coherence and harnessing single-pixel imaging technique, we demonstrate interaction-free, single-pixel quantum imaging of a structured object with undetected photons. Thereby, we push the capability of quantum imaging to the extreme point in which no interaction is required between object and photons and the detection requirement is greatly reduced. Our work paves the path for applications in characterizing delicate samples with single-pixel imaging at silicon-detectable wavelengths.

Fisher information and the Cramér-Rao lower bound in single-pixel localization microscopy with spatiotemporally modulated illumination.

Single-pixel imaging, the concept that an image can be captured via a single-pixel detector, is a cost-effective yet powerful technique to reduce data acquisition duration without sacrificing image resolution when properly structured illumination patterns are introduced. Normally, the image reconstruction process is subject to the diffraction limit. Here, we study the possibility of exploiting the information contained in the illumination patterns to enable a form of single-pixel localization microscopy (SPLM) for super-resolution. This concept is inspired by coherent holographic image reconstruction by phase transfer (CHIRPT) microscopy. CHIRPT microscopy is a single-pixel imaging technique that uses structured illumination that is spatiotemporally modulated (STM) so that a unique temporal modulation pattern is imparted to each point within a large illumination volume. The fluorescent light emitted by molecules contains the same temporal modulations as the illumination patterns at the locations of the molecules. By recording a portion of the total emitted fluorescent power, the signal may be numerically processed to form an image. Unique temporal modulation patterns that excite fluorescent probes at each point can also be used to localize individual molecules by matching their particular temporal light emission patterns to the measured temporal signal. This paper evaluates the feasibility of SPLM with STM illuminations used in and inspired by CHIRPT microscopy via the information content its data carry about the emitter location(s). More specifically, we provide the mathematical formalism of Fisher information (FI) and the Cramér-Rao lower bound (CRLB) associated with the location parameters of the emitter(s). The FI and CRLB are then numerically evaluated under different experimental assumptions to assess the effects of experimental parameters on localization precision. Last, we compare the single-pixel CRLB to that from camera-based single-molecule localization microscopy in the localization of a single emitter. We show that SPLM has several distinguishing characteristics that provide certain advantages, such as relatively constant CRLB over a very large illumination volume and improved CRLB for 3D localization due to the information coupling introduced by simultaneous modulations of the transverse axes.

Single-pixel imaging using discrete Zernike moments.

A novel single-pixel imaging (SPI) technique based on discrete orthogonal Zernike moments is proposed. In this technique, the target object is illuminated by two sets of Zernike basis patterns which satisfy the Zernike polynomials. The Zernike moments of object image are obtained by measuring the reflected light intensities through a single-pixel detector. And the object image is reconstructed by summing the product of Zernike polynomials and detected intensities iteratively. By theoretical and experimental demonstrations, an image with high quality is retrieved under compressive sampling. Moreover, the Zernike illuminating patterns are used for object classification due to the rotation invariant of Zernike moments. By measuring the amplitudes of a few specific Zernike moments through the SPI system, the rotated images with different angles and the same content are classified into the same class on experiment. This classification technique has the advantages of high efficiency and high accuracy due to the high modulation speed and high sensitivity of SPI system.

Imaging of Single-Photon Orbital-Angular-Momentum Bell States

Bell states are the most fundamental resource in realizing quantum information tasks and have a very unique position in quantum mechanics. Utilizing orbital angular momentum (OAM) to encode single-photon Bell states enables the realization of high-dimensional Hilbert space, which is crucial for the field of quantum information. In this paper, we design a single-photon OAM Bell-state evolution device based on the Sagnac interferometer, which can make one-to-one correspondence between the input Bell states and the output states. Moreover, we also develop a single-photon single-pixel imaging (SPI) technique, which improves spatial resolution and decreases acquisition time in the meantime, to capture interference images of output states. The results show that single-photon OAM Bell states can be fully recognized by comparing the differences of interference images. We innovatively utilize the SPI technique for single-photon OAM Bell-state recognition. This indicates that the SPI technique effectively promotes the study of quantum information based on OAM, while the quantum information based on OAM provides a clear application scenario for the SPI technique.