Mark W. Koch

A 2D Range Hausdorff Approach for 3D Face Recognition

This paper presents a 3D facial recognition algorithm based on the Hausdorff distance metric. The standard 3D formulation of the Hausdorff matching algorithm has been modified to operate on a 2D range image, enabling a reduction in computation from O(N2) to O(N) without large storage requirements. The Hausdorff distance is known for its robustness to data outliers and inconsistent data between two data sets, making it a suitable choice for dealing with the inherent problems in many 3D datasets due to sensor noise and object self-occlusion. For optimal performance, the algorithm assumes a good initial alignment between probe and template datasets. However, to minimize the error between two faces, the alignment can be iteratively refined. Results from the algorithm are presented using 3D face images from the Face Recognition Grand Challenge database version 1.0.

Cueing, feature discovery, and one-class learning for synthetic aperture radar automatic target recognition

Abstract The exquisite capabilities of biological neural systems for recognizing target patterns subject to large variations have motivated us to investigate neurophysiologically-inspired techniques for automatic target recognition. This paper describes a modular multi-stage architecture for focus-of-attention cueing, feature discovery and extraction, and one-class pattern learning and identification in synthetic aperture radar imagery. To prescreen massive amounts of image data, we apply a focus-of-attention algorithm using data skewness to extract man-made objects from natural clutter regions. We apply self-organizing feature discovery algorithms that uniquely characterize targets in a reduced dimension space and use self-organizing one-class classifiers for learning target variations. We also develop a distance metric for partial obscuration recognition. We present performance results using simulated SAR data and test for within-class generalization using nontrained targets including both in-the-clear and partially obscured examples. We test for between-class generalization using non-trained targets including both in-the-clear and partially obscured examples. We test for between-class generalization using near-target data.

Distributed Sensor Fusion in Water Quality Event Detection

To protect drinking water systems, a contamination warning system can use in-line sensors to indicate possible accidental and deliberate contamination. Currently, reporting of an incident occurs when data from a single station detects an anomaly. This paper proposes an approach for combining data from multiple stations to reduce false background alarms. By considering the location and time of individual detections as points resulting from a random space-time point process, Kulldorff’s scan test can find statistically significant clusters of detections. Using EPANET to simulate contaminant plumes of varying sizes moving through a water network with varying amounts of sensing nodes, it is shown that the scan test can detect significant clusters of events. Also, these significant clusters can reduce the false alarms resulting from background noise and the clusters can help indicate the time and source location of the contaminant. Fusion of monitoring station results within a moderately sized network show fal...

3D facial recognition: a quantitative analysis

Two-dimensional facial recognition has, traditionally, been an attractive biometric, however, the accuracy of 2D facial recognition (FR) is performance limited and insufficient when confronted with extensive numbers of people to screen and identify, and the numerous appearances that a 2D face can exhibit. In efforts to overcome many of the issues limiting 2D FR technology, researchers are beginning to focus their attention on 3D FR technology. In this paper, an analysis of a 3D FR system being developed at Sandia National Laboratories is performed. The study involves the use of 200 subjects on which verification (one-to-one) matches are performed using a single probe database (one correct match per subject) and 30 subjects on which identification matches are performed. The system is evaluated in terms of probability of detection (Pd) and probability of false accepts (FAR). The results presented will aid in providing an initial understanding of the performance of 3D FR.

A Sequential Vehicle Classifier for Infrared Video using Multinomial Pattern Matching

Vehicle classification is a challenging problem, since vehicles can take on many different appearances and sizes due to their form and function, and the viewing conditions. The low resolution of uncooled-infrared video and the large variability of naturally occurring environmental conditions can make this an even more difficult problem. We develop a multilook fusion approach for improving the performance of a single look system. Our single look approach is based on extracting a signature consisting of a histogram of gradient orientations from a set of regions covering the moving object. We use the multinomial pattern matching algorithm to match the signature to a database of learned signatures. To combine the match scores of multiple signatures from a single tracked object, we use the sequential probability ratio test. Using real infrared data we show excellent classification performance, with low expected error rates, when using at least 25 looks.

Feature discovery via neural networks for object recognition in SAR imagery

A two-stage self-organizing neural network architecture has been applied to object recognition in synthetic aperture radar imagery. The first stage performs feature extraction and implements a two-layer neocognitron. The resulting feature vectors are presented to the second stage, an ART 2-A classifier network, which clusters the features into multiple target categories. Training is performed off-line in two steps. First, the neocognitron self-organizes in response to repeated presentations of an object to recognize. During this training process, discovered features and the mechanisms for their extraction are captured in the excitatory weight patterns. In the second step, neocognitron learning is inhibited and the ART 2-A classifier forms categories in response to the feature vectors generated by additional presentations of the object to recognize. Finally, all training is inhibited and the system tested against a variety of objects and background clutter. The results of the initial experiments are reported.<<ETX>>

Multinomial pattern matching for high range resolution radar profiles

Airborne ground moving-target indication (GMTI) radar can track moving vehicles at large standoff distances. Unfortunately, trajectories from multiple vehicles can become kinematically ambiguous, resulting in confusion between a target vehicle of interest and other vehicles. We propose the use of high range resolution (HRR) radar profiles and multinomial pattern matching (MPM) for target fingerprinting and track stitching to overcome kinematic ambiguities. Sandias MPM algorithm is a robust template-based identification algorithm that has been applied successfully to various target recognition problems. MPM utilizes a quantile transformation to map target intensity samples to a small number of grayscale values, or quantiles. The algorithm relies on a statistical characterization of the multinomial distribution of the sample-by-sample intensity values for target profiles. The quantile transformation and statistical characterization procedures are extremely well suited to a robust representation of targets for HRR profiles: they are invariant to sensor calibration, robust to target signature variations, and lend themselves to efficient matching algorithms. In typical HRR tracking applications, target fingerprints must be initiated on the fly from a limited number of HRR profiles. Data may accumulate indefinitely as vehicles are tracked, and their templates must be continually updated without becoming unbounded in size or complexity. To address this need, an incrementally updated version of MPM has been developed. This implementation of MPM incorporates individual HRR profiles as they become available, and fuses data from multiple aspect angles for a given target to aid in track stitching. This paper provides a description of the incrementally updated version of MPM.

Classifying Acoustic Signatures Using the Sequential Probability Ratio Test

Abstract Acoustic sensors can provide real time information about moving targets. The acoustic information is typically processed sequentially, allowing the sequential probability ratio test (SPRT) to be used as the basis to solve the target identification problem. The SPRT keeps gathering observations only as long as the statistical test has a value between the upper stopping boundary and the lower stopping boundary. When the test goes above the upper boundary or below the lower boundary, the system can make a decision. The desired false alarm error rate and the desired missed detection error rate determine the upper and lower stopping boundaries. We present extensions to the sequential probability ratio test to handle problems of dependence, contamination, and the unknown class. We also present results for using the SPRT for target identification using acoustic information.

Feature discovery in gray level imagery for one-class object recognition

Feature extraction transforms an objects image representation to an alternate reduced representation. Feature selection can be time-consuming and difficult to optimize so we have investigated unsupervised neural networks for feature discovery. We first discuss an inherent limitation in competitive type neural networks for discovering features in gray level images. We then show how Sangers Generalized Hebbian Algorithm (GHA) removes this limitation and describe a novel GHA application for learning object features that discriminate the object from clutter. Using a specific example, we show how these features are better at distinguishing the target object from other nontarget objects with Carpenters ART 2-A as the pattern classifier.<<ETX>>

Vehicle Classification in Infrared Video Using the Sequential Probability Ratio Test

This chapter develops a multilook fusion approach for improving the performance of a single-look vehicle classification system for infrared video. Vehicle classification is a challenging problem since vehicles can take on many different appearances and sizes due to their form and function and the viewing conditions. The low resolution of uncooled infrared video and the large variability of naturally occurring environmental conditions can make this an even more difficult problem. Our single-look approach is based on extracting a signature consisting of a histogram of gradient orientations from a set of regions covering the moving object. We use the multinomial pattern-matching algorithm to match the signature to a database of learned signatures. To combine the match scores of multiple signatures from a single tracked object, we use the sequential probability ratio test. Using infrared data, we show excellent classification performance, with low expected error rates, when using at least 25 looks.