CMOS approach to compressed-domain image acquisition.

A hardware implementation of a real-time compressed-domain image acquisition system is demonstrated. The system performs front-end computational imaging, whereby the inner product between an image and an arbitrarily-specified mask is implemented in silicon. The acquisition system is based on an intelligent readout integrated circuit (iROIC) that is capable of providing independent bias voltages to individual detectors, which enables implementation of spatial multiplication with any prescribed mask through a bias-controlled response-modulation mechanism. The modulated pixels are summed up in the image grabber to generate the compressed samples, namely aperture-coded coefficients, of an image. A rigorous bias-selection algorithm is presented to the readout circuit, which exploits the bias-dependent nature of the imager's responsivity. Proven functionality of the hardware in transform coding compressed image acquisition, silicon-level compressive sampling, in pixel nonuniformity correction and hardware-level implementation of region-based enhancement is demonstrated.

On the analytical formulation of excess noise in avalanche photodiodes with dead space.

Simple, approximate formulas are developed to calculate the mean gain and excess noise factor for avalanche photodiodes using the dead-space multiplication theory in the regime of small multiplication width and high applied electric field. The accuracy of the approximation is investigated by comparing it to the exact numerical method using recursive coupled integral equations and it is found that it works for dead spaces up to 15% of the multiplication width, which is substantial. The approximation is also tested for real materials such as GaAs, InP and Si for various multiplication widths, and the results found are accurate within ∼ 15% of the actual noise, which is a significant improvement over the local-theory noise formula. The results obtained for the mean gain also confirm the recently reported relationship between experimentally determined local ionization coefficients and the enabled non-local ionization coefficients.

Breaking the buildup-time limit of sensitivity in avalanche photodiodes by dynamic biasing.

Avalanche photodiodes (APDs) are the preferred photodetectors for direct-detection, high data-rate long-haul optical telecommunications. APDs can detect low-level optical signals due to their internal amplification of the photon-generated electrical current, which is attributable to the avalanche of electron and hole impact ionizations. Despite recent advances in APDs aimed at reducing the average avalanche-buildup time, which causes intersymbol interference and compromises receiver sensitivity at high data rates, operable speeds of commercially available APDs have been limited to 10Gbps. We report the first demonstration of a dynamically biased APD that breaks the traditional sensitivity-versus-speed limit by employing a data-synchronous sinusoidal reverse-bias that drastically suppresses the average avalanche-buildup time. Compared with traditional DC biasing, the sensitivity of germanium APDs at 3Gbps is improved by 4.3 dB, which is equivalent to a 3,500-fold reduction in the bit-error rate. The method is APD-type agnostic and it promises to enable operation at rates of 25Gbps and beyond.

Thin 3D multiplication regions in plasmonically enhanced nanopillar avalanche detectors.

We demonstrate a nanopillar (NP) device structure for implementing plasmonically enhanced avalanche photodetector arrays with thin avalanche volumes (∼ 310 nm × 150 nm × 150 nm). A localized 3D electric field due to a core-shell PN junction in a NP acts as a multiplication region, while efficient light absorption takes place via surface plasmon polariton Bloch wave (SPP-BW) modes due to a self-aligned metal nanohole lattice. Avalanche gains of ∼216 at 730 nm at -12 V are obtained. We show through capacitance-voltage characterization, temperature-dependent breakdown measurements, and detailed device modeling that the avalanche region is on the order of the ionization path length, such that dead-space effects become significant. This work presents a clear path toward engineering dead space effects in thin 3D-confined multiplication regions for high performance avalanche detectors for applications in telecommunications, sensing and single photon detection.

Multiplication theory for dynamically biased avalanche photodiodes: new limits for gain bandwidth product.

Novel theory is developed for the avalanche multiplication process in avalanche photodiodes (APDs) under time-varying reverse-biasing conditions. Integral equations are derived characterizing the statistics of the multiplication factor and the impulse-response function of APDs, as well as their breakdown probability, all under the assumption that the electric field driving the avalanche process is time varying and spatially nonuniform. Numerical calculations generated by the model predict that by using a bit-synchronous sinusoidal biasing scheme to operate the APD in an optical receiver, the pulse-integrated gain-bandwidth product can be improved by a factor of 5 compared to the same APD operating under the conventional static biasing. The bit-synchronized periodic modulation of the electric field in the multiplication region serves to (1) produce large avalanche multiplication factors with suppressed avalanche durations for photons arriving in the early phase of each optical pulse; and (2) generate low avalanche gains and very short avalanche durations for photons arriving in the latter part of each optical pulse. These two factors can work together to reduce intersymbol interference in optical receivers without sacrificing sensitivity.

Continuous time-varying biasing approach for spectrally tunable infrared detectors.

In a recently demonstrated algorithmic spectral-tuning technique by Jang et al. [Opt. Express 19, 19454-19472, (2011)], the reconstruction of an object's emissivity at an arbitrarily specified spectral window of interest in the long-wave infrared region was achieved. The technique relied upon forming a weighted superposition of a series of photocurrents from a quantum dots-in-a-well (DWELL) photodetector operated at discrete static biases that were applied serially. Here, the technique is generalized such that a continuously varying biasing voltage is employed over an extended acquisition time, in place using a series of fixed biases over each sub-acquisition time, which totally eliminates the need for the post-processing step comprising the weighted superposition of the discrete photocurrents. To enable this capability, an algorithm is developed for designing the time-varying bias for an arbitrary spectral-sensing window of interest. Since continuous-time biasing can be implemented within the readout circuit of a focal-plane array, this generalization would pave the way for the implementation of the algorithmic spectral tuning in focal-plane arrays within in each frame time without the need for on-sensor multiplications and additions. The technique is validated by means of simulations in the context of spectrometry and object classification while using experimental data for the DWELL under realistic signal-to-noise ratios.

New perspective on passively quenched single photon avalanche diodes: effect of feedback on impact ionization.

Single-photon avalanche diodes (SPADs) are primary devices in photon counting systems used in quantum cryptography, time resolved spectroscopy and photon counting optical communication. SPADs convert each photo-generated electron hole pair to a measurable current via an avalanche of impact ionizations. In this paper, a stochastically self-regulating avalanche model for passively quenched SPADs is presented. The model predicts, in qualitative agreement with experiments, three important phenomena that traditional models are unable to predict. These are: (1) an oscillatory behavior of the persistent avalanche current; (2) an exponential (memoryless) decay of the probability density function of the stochastic quenching time of the persistent avalanche current; and (3) a fast collapse of the avalanche current, under strong feedback conditions, preventing the development of a persistent avalanche current. The model specifically captures the effect of the load's feedback on the stochastic avalanche multiplication, an effect believed to be key in breaking today's counting rate barrier in the 1.55-µm detection window.

Information content per photon versus image fidelity in three-dimensional photon-counting integral imaging.

Photon-counting integral imaging has been introduced recently, and its applications in three-dimensional (3D) object sensing, visualization, recognition, and classification under photon-starved conditions have been demonstrated. This paper sheds light on the underlying information-theoretic foundation behind the ability of photon-counting integral imaging in performing complex tasks with far fewer photons than conventional imaging systems. A metric for photon-information content is formulated in the context of 3D photon-counting imaging, and its properties are investigated. It is shown that there is an inherent trade-off between imaging fidelity, measured by the entropy-normalized mutual information associated with a given imaging system, and the amount of information in each photon used in the imaging process, as represented by the photon-number-normalized mutual information. The dependence of this trade-off on photon statistics, correlation in the 3D image, and the signal-to-noise ratio of the photon-detection system is also investigated.

Data compressive paradigm for multispectral sensing using tunable DWELL mid-infrared detectors.

While quantum dots-in-a-well (DWELL) infrared photodetectors have the feature that their spectral responses can be shifted continuously by varying the applied bias, the width of the spectral response at any applied bias is not sufficiently narrow for use in multispectral sensing without the aid of spectral filters. To achieve higher spectral resolutions without using physical spectral filters, algorithms have been developed for post-processing the DWELL's bias-dependent photocurrents resulting from probing an object of interest repeatedly over a wide range of applied biases. At the heart of these algorithms is the ability to approximate an arbitrary spectral filter, which we desire the DWELL-algorithm combination to mimic, by forming a weighted superposition of the DWELL's non-orthogonal spectral responses over a range of applied biases. However, these algorithms assume availability of abundant DWELL data over a large number of applied biases (>30), leading to large overall acquisition times in proportion with the number of biases. This paper reports a new multispectral sensing algorithm to substantially compress the number of necessary bias values subject to a prescribed performance level across multiple sensing applications. The algorithm identifies a minimal set of biases to be used in sensing only the relevant spectral information for remote-sensing applications of interest. Experimental results on target spectrometry and classification demonstrate a reduction in the number of required biases by a factor of 7 (e.g., from 30 to 4). The tradeoff between performance and bias compression is thoroughly investigated.

Information theoretic approach for assessing image fidelity in photon-counting arrays.

The method of photon-counting integral imaging has been introduced recently for three-dimensional object sensing, visualization, recognition and classification of scenes under photon-starved conditions. This paper presents an information-theoretic model for the photon-counting imaging (PCI) method, thereby providing a rigorous foundation for the merits of PCI in terms of image fidelity. This, in turn, can facilitate our understanding of the demonstrated success of photon-counting integral imaging in compressive imaging and classification. The mutual information between the source and photon-counted images is derived in a Markov random field setting and normalized by the source-image's entropy, yielding a fidelity metric that is between zero and unity, which respectively corresponds to complete loss of information and full preservation of information. Calculations suggest that the PCI fidelity metric increases with spatial correlation in source image, from which we infer that the PCI method is particularly effective for source images with high spatial correlation; the metric also increases with the reduction in photon-number uncertainty. As an application to the theory, an image-classification problem is considered showing a congruous relationship between the fidelity metric and classifier's performance.

Detection theory for accurate and non-invasive skin cancer diagnosis using dynamic thermal imaging.

Skin cancer is the most common cancer in the United States with over 3.5M annual cases. Presently, visual inspection by a dermatologist has good sensitivity (> 90%) but poor specificity (< 10%), especially for melanoma, which leads to a high number of unnecessary biopsies. Here we use dynamic thermal imaging (DTI) to demonstrate a rapid, accurate and non-invasive imaging system for detection of skin cancer. In DTI, the lesion is cooled down and the thermal recovery is recorded using infrared imaging. The thermal recovery curves of the suspected lesions are then utilized in the context of continuous-time detection theory in order to define an optimal statistical decision rule such that the sensitivity of the algorithm is guaranteed to be at a maximum for every prescribed false-alarm probability. The proposed methodology was tested in a pilot study including 140 human subjects demonstrating a sensitivity in excess of 99% for a prescribed specificity in excess of 99% for detection of skin cancer. To the best of our knowledge, this is the highest reported accuracy for any non-invasive skin cancer diagnosis method.

Plasmonic field confinement for separate absorption-multiplication in InGaAs nanopillar avalanche photodiodes.

Avalanche photodiodes (APDs) are essential components in quantum key distribution systems and active imaging systems requiring both ultrafast response time to measure photon time of flight and high gain to detect low photon flux. The internal gain of an APD can improve system signal-to-noise ratio (SNR). Excess noise is typically kept low through the selection of material with intrinsically low excess noise, using separate-absorption-multiplication (SAM) heterostructures, or taking advantage of the dead-space effect using thin multiplication regions. In this work we demonstrate the first measurement of excess noise and gain-bandwidth product in III-V nanopillars exhibiting substantially lower excess noise factors compared to bulk and gain-bandwidth products greater than 200 GHz. The nanopillar optical antenna avalanche detector (NOAAD) architecture is utilized for spatially separating the absorption region from the avalanche region via the NOA resulting in single carrier injection without the use of a traditional SAM heterostructure.

Maximum-likelihood image estimation using photon-correlated beams.

A theory is presented addressing the fundamental limits of image estimation in a setup that uses two photon-correlated beams. These beams have the property that their photon arrivals, as a point process, are ideally synchronized in time and space. The true image represents the spatial distribution of the optical transmittance (or reflectance) of an object. In this setup, one beam is used to probe the image while the other is used as a reference providing additional information on the actual number of photons impinging on the object. This additional information is exploited to reduce the effect of quantum noise associated with the uncertainty in the number of photons per pixel. A stochastic model for the joint statistics of the two observation matrices is developed and used to obtain a local maximum-likelihood estimator of the image. The model captures the nonideal nature of the correlation between the photons of the beams by means of a simple random translation model. The mean-square error of the estimator is evaluated and compared to the corresponding conventional techniques. Conditions for the performance advantage of the proposed estimator are examined in terms of key system parameters. The theoretical predictions are demonstrated by means of simulation.

Model-based edge detector for spectral imagery using sparse spatiospectral masks.

Two model-based algorithms for edge detection in spectral imagery are developed that specifically target capturing intrinsic features such as isoluminant edges that are characterized by a jump in color but not in intensity. Given prior knowledge of the classes of reflectance or emittance spectra associated with candidate objects in a scene, a small set of spectral-band ratios, which most profoundly identify the edge between each pair of materials, are selected to define a edge signature. The bands that form the edge signature are fed into a spatial mask, producing a sparse joint spatiospectral nonlinear operator. The first algorithm achieves edge detection for every material pair by matching the response of the operator at every pixel with the edge signature for the pair of materials. The second algorithm is a classifier-enhanced extension of the first algorithm that adaptively accentuates distinctive features before applying the spatiospectral operator. Both algorithms are extensively verified using spectral imagery from the airborne hyperspectral imager and from a dots-in-a-well midinfrared imager. In both cases, the multicolor gradient (MCG) and the hyperspectral/spatial detection of edges (HySPADE) edge detectors are used as a benchmark for comparison. The results demonstrate that the proposed algorithms outperform the MCG and HySPADE edge detectors in accuracy, especially when isoluminant edges are present. By requiring only a few bands as input to the spatiospectral operator, the algorithms enable significant levels of data compression in band selection. In the presented examples, the required operations per pixel are reduced by a factor of 71 with respect to those required by the MCG edge detector.

Generalized signal-to-noise ratio for spectral sensors with correlated bands.

Traditional spectral sensors are intentionally designed to minimize overlap among spectral response functions of different bands. In contrast, some emerging classes of spectral sensors exhibit significant band overlap. An effect introduced by such band overlap is that the photodetector noise of one band is coupled into the others in subsequent data processing steps. Because of this, the traditional band-by-band definition of signal-to-noise ratio (SNR) cannot fully describe the detector's noise level. We devise a general definition of SNR in spectral space based on a recently developed geometrical spectral imaging model [J. Opt. Soc. Am. A24, 2864 (2007)]. With this model, we can find an orthogonal basis of the spectral response functions for the spectral sensor with decreasing instrument SNRs. We can also define the average instrument SNR for the whole sensor, which makes it possible to characterize quantitatively the photodetector noise of a spectral sensor with correlated bands.

Data interpretation for spectral sensors with correlated bands.

New classes of spectral sensors are emerging that have significant overlap in the band spectral response functions. While conventional sensors such as the Multispectral Thermal Images (MTI) or Landsat may have responses with a few percent overlap between adjacent bands, some of the emerging sensors can have more than 50% correlation among all spectral bands. The traditional geometrical models used to describe spectral data fail when such high levels of correlation exist. In this paper we present a generalized geometrical model that relies on functional analysis. We define a sensor space and a scene space that can be used to characterize the suitability of a sensor for a particular spectral sensing task. We demonstrate that classifiers based on first-order distance and angle metrics fail for sensors with highly correlated bands unless appropriate preprocessing is carried out. We further show that second-order statistical classifiers are largely immune to many of the problems introduced by the correlated band responses.

Scene estimation from speckled synthetic aperture radar imagery: Markov-random-field approach.

A novel Markov-random-field model for speckled synthetic aperture radar (SAR) imagery is derived according to the physical, spatial statistical properties of speckle noise in coherent imaging. A convex Gibbs energy function for speckled images is derived and utilized to perform speckle-compensating image estimation. The image estimation is formed by computing the conditional expectation of the noisy image at each pixel given its neighbors, which is further expressed in terms of the derived Gibbs energy function. The efficacy of the proposed technique, in terms of reducing speckle noise while preserving spatial resolution, is studied by using both real and simulated SAR imagery. Using a number of commonly used metrics, the performance of the proposed technique is shown to surpass that of existing speckle-noise-filtering methods such as the Gamma MAP, the modified Lee, and the enhanced Frost.

Multimodel Kalman filtering for adaptive nonuniformity correction in infrared sensors.

We present an adaptive technique for the estimation of nonuniformity parameters of infrared focal-plane arrays that is robust with respect to changes and uncertainties in scene and sensor characteristics. The proposed algorithm is based on using a bank of Kalman filters in parallel. Each filter independently estimates state variables comprising the gain and the bias matrices of the sensor, according to its own dynamic-model parameters. The supervising component of the algorithm then generates the final estimates of the state variables by forming a weighted superposition of all the estimates rendered by each Kalman filter. The weights are computed and updated iteratively, according to the a posteriori-likelihood principle. The performance of the estimator and its ability to compensate for fixed-pattern noise is tested using both simulated and real data obtained from two cameras operating in the mid- and long-wave infrared regime.

Statistical adaptive sensing by detectors with spectrally overlapping bands.

It has recently been reported that by using a spectral-tuning algorithm, the photocurrents of multiple detectors with spectrally overlapping responsivities can be optimally combined to synthesize, within certain limits, the response of a detector with an arbitrary responsivity. However, it is known that the presence of noise in the photocurrent can degrade the performance of this algorithm significantly, depending on the choice of the responsivity spectrum to be synthesized. We generalize this algorithm to accommodate noise. The results are applied to quantum-dot mid-infrared detectors with bias-dependent spectral responses. Simulation and experiment are used to show the ability of the algorithm to reduce the adverse effect of noise on its spectral-tuning capability.

Generalized algebraic scene-based nonuniformity correction algorithm.

A generalization of a recently developed algebraic scene-based nonuniformity correction algorithm for focal plane array (FPA) sensors is presented. The new technique uses pairs of image frames exhibiting arbitrary one- or two-dimensional translational motion to compute compensator quantities that are then used to remove nonuniformity in the bias of the FPA response. Unlike its predecessor, the generalization does not require the use of either a blackbody calibration target or a shutter. The algorithm has a low computational overhead, lending itself to real-time hardware implementation. The high-quality correction ability of this technique is demonstrated through application to real IR data from both cooled and uncooled infrared FPAs. A theoretical and experimental error analysis is performed to study the accuracy of the bias compensator estimates in the presence of two main sources of error.