K. P. Judd

Characterization of the AWARE 10 two-gigapixel wide-field-of-view visible imager

System requirements for many military electro-optic and IR camera systems reflect the need for both wide-field-of-view situational awareness as well as high-resolution imaging for target identification. In this work we present a new imaging system architecture designed to perform both functions simultaneously and the AWARE 10 camera as an example at visible wavelengths. We first describe the basic system architecture and user interface followed by a laboratory characterization of the system optical performance. We then describe a field experiment in which the camera was used to identify several maritime targets at varying range. The experimental results indicate that users of the system are able to correctly identify ~10 m targets at between 4 and 6 km with 70% accuracy.

Estimating detection and identification probabilities in maritime target acquisition

This work describes several approaches to the estimation of target detection and identification probabilities as a function of target range. A Bayesian approach to estimation is adopted, whereby the posterior probability distributions associated with these probabilities are analytically derived. The parameter posteriors are then used to develop credible intervals quantifying the degree of uncertainty in the parameter estimates. In our first approach we simply show how these credible intervals evolve as a function of range. A second approach, also following the Bayesian philosophy, attempts to directly estimate the parameterized performance curves. This second approach makes efficient use of the available data and yields a distribution of probability versus range curves. Finally, we demonstrate both approaches using experimental data collected from wide field-of-view imagers focused on maritime targets.

Manifold learning techniques for unsupervised anomaly detection

Abstract Appropriately identifying outlier data is a critical requirement in the decision-making process of many expert and intelligent systems deployed in a variety of fields including finance, medicine, and defense. Classical outlier detection schemes typically rely on the assumption that normal/background data of interest are distributed according to an assumed statistical model and search for data that deviate from that assumption. However, it is frequently the case that performance is reduced because the underlying distribution does not follow the assumed model. Manifold learning techniques offer improved performance by learning better models of the background but can be too computationally expensive due to the need to calculate a distance measure between all data points. Here, we study a general framework that allows manifold learning techniques to be used for unsupervised anomaly detection by reducing computational expense via a uniform random sampling of a small fraction of the data. A background manifold is learned from the sample and then an out-of-sample extension is used to project unsampled data into the learned manifold space and construct an anomaly detection statistic based on the prediction error of the learned manifold. The method works well for unsupervised anomaly detection because, by definition, the ratio of anomalous to non-anomalous data points is small and the sampling will be dominated by background points. However, a variety of parameters that affect detection performance are introduced so we use here a low-dimensional toy problem to investigate their effect on the performance of four learning algorithms (kernel PCA, two versions of diffusion map, and the Parzen density estimator). We then apply the methods to the detection of watercraft in an ensemble of 22 infrared maritime scenes where we find kernel PCA to be superior and show that it outperforms a commonly employed baseline algorithm. The framework is not limited to the tested image processing example and can be used for any unsupervised anomaly detection task.

Experimental and numerical study of underwater beam propagation in a Rayleigh–Bénard turbulence tank

The propagation of a laser beam through Rayleigh-Bénard (RB) turbulence is investigated experimentally and by way of numerical simulation. For the experimental part, a focused laser beam transversed a 5  m×0.5  m×0.5  m water filled tank lengthwise. The tank is heated from the bottom and cooled from the top to produce convective RB turbulence. The effect of the turbulence on the beam is recorded on the exit of the beam from the tank. From the centroid motion of the beam, the index of refraction structure constant Cn2 is determined. For the numerical efforts RB turbulence is simulated for a tank of the same geometry. The simulated temperature fields are converted to the index of refraction distributions, and Cn2 is extracted from the index of refraction structure functions, as well as from the simulated beam wander. To model the effect on beam propagation, the simulated index of refraction fields are converted to discrete index of refraction phase screens. These phase screens are then used in a split-step beam propagation method to investigate the effect of the turbulence on a laser beam. The beam wander as well as the index of refraction structure parameter Cn2 determined from the experiment and simulation are compared and found to be in good agreement.

Range performance of the DARPA AWARE wide field-of-view visible imager

In a prior paper, we described a new imaging architecture that addresses the need for wide field-of-view imaging combined with the resolution required to identify targets at long range. Over the last two years substantive improvements have been made to the system, both in terms of the size, weight, and power of the camera as well as to the optics and data management software. The result is an overall improvement in system performance, which we demonstrate via a maritime target identification experiment.

Improving sparse representation algorithms for maritime video processing

We present several improvements to published algorithms for sparse image modeling with the goal of improving processing of imagery of small watercraft in littoral environments. The first improvement is to the K-SVD algorithm for training over-complete dictionaries, which are used in sparse representations. It is shown that the training converges significantly faster by incorporating multiple dictionary (i.e., codebook) update stages in each training iteration. The paper also provides several useful and practical lessons learned from our experience with sparse representations. Results of three applications of sparse representation are presented and compared to the state-of-the-art methods; image compression, image denoising, and super-resolution.

Watercraft detection in short-wave infrared imagery using a tailored wavelet basis

We present a technique for small watercraft detection in a littoral environment characterized by multiple targets and both land- and sea-based clutter. The detector correlates a tailored wavelet model trained from previous imagery with newly acquired scenes. An optimization routine is used to learn a wavelet signal model that improves the average probability of detection for a xed false alarm rate on an ensemble of training images. The resulting wavelet is shown to improve detection on a previously unseen set of test images. Performance is quantied with ROC curves.

Determining seeing conditions of a horizontal turbulent optical path with video image analysis

The turbulent effects from the Earth’s atmosphere degrade the performance of any optical system within it. There have been numerous studies in the effects of atmospheric turbulence on an imaging system that is pointed vertically to the sky looking at distant objects and the seeing conditions associated with it. We investigate the calculation of the seeing conditions with an imaging system pointed horizontally in terrestrial and maritime environments. We have acquired video data of different horizontal paths in the infrared wavelengths and performed data analysis that will be the basis of new characterizations and modeling of horizontal path atmospheric turbulence.

Discriminative dictionaries for automated target recognition

We present an approach for discriminating among dierent classes of imagery in a scene. Our intended application is the detection of small watercraft in a littoral environment where both targets and land- and sea-based clutter are present. The approach works by training dierent overcomplete dictionaries to model the dierent image classes. The likelihood ratio obtained by applying each model to the unknown image is then used as the discriminating test statistic. We rst demonstrate the approach on an illustrative test problem and then apply the algorithm to short-wave infrared imagery with known targets.