Dr. Gerald J. Dobeck Profile

Dr. Gerald J. Dobeck

Research Engineer at Naval Surface Warfare Ctr

SPIE Involvement:

Author

Publications (47)

Proceedings Article | 29 April 2010 Paper

Adaptive large-scale clutter removal from imagery with application to high-resolution sonar imagery

Gerald Dobeck

Proceedings Volume 7664, 76640X (2010) https://doi.org/10.1117/12.851037

KEYWORDS: Denoising, Detection and tracking algorithms, Target detection, Spatial frequencies, Synthetic aperture radar, Reflectivity, Fourier transforms, Image processing, Bismuth, Image segmentation

Read Abstract +

Proceedings Article | 26 April 2007 Paper

Coherent-based method for detection of underwater objects from sonar imagery

Mahmood Azimi-Sadjadi, Gerry Dobeck, James Tucker

Proceedings Volume 6553, 65530U (2007) https://doi.org/10.1117/12.718837

KEYWORDS: Target detection, Feature extraction, Simulation of CCA and DLA aggregates, Sensors, Submerged target detection, Image classification, Databases, Image filtering, Warfare, Canonical correlation analysis

Read Abstract +

Proceedings Article | 9 April 2007 Paper

A weighted quadratic asymptotic analysis of cost functions used in classifier design with extensions to finite-size training sets

Gerald Dobeck

Proceedings Volume 6576, 657611 (2007) https://doi.org/10.1117/12.720614

KEYWORDS: Brain-machine interfaces, Probability theory, Statistical analysis, Inspection, Warfare, Bridges, Neural networks, Algorithm development, Network architectures, Independent component analysis

Read Abstract +

Proceedings Article | 10 June 2005 Paper

Volterra fusion of processing strings for automated sea mine classification in shallow water

Manuel Fernandez, Gerald Dobeck, Tom Aridgides

Proceedings Volume 5794, (2005) https://doi.org/10.1117/12.602116

KEYWORDS: Detection and tracking algorithms, Image processing, Algorithm development, Feature extraction, Mining, Digital filtering, Naval mines, Data fusion, Image fusion, Feature selection

Read Abstract +

Proceedings Article | 10 June 2005 Paper

Image normalization using the serpentine forward-backward filter: application to high-resolution sonar imagery and its impact on mine detection and classification

Gerald Dobeck

Proceedings Volume 5794, (2005) https://doi.org/10.1117/12.602878

KEYWORDS: Image filtering, Digital filtering, Land mines, Image segmentation, Image processing, Linear filtering, Electronic filtering, Image classification, Automatic target recognition, Acoustics

Read Abstract +

Proceedings Article | 21 September 2004 Paper

Improved processing-string fusion-approach investigation for automated sea-mine classification in shallow water

Gerald Dobeck, Tom Aridgides, Manuel Fernandez

Proceedings Volume 5415, (2004) https://doi.org/10.1117/12.541007

KEYWORDS: Detection and tracking algorithms, Image processing, Mining, Feature extraction, Data fusion, Digital filtering, Algorithm development, Image fusion, Naval mines, Image classification

Read Abstract +

An improved sea mine computer-aided-detection/computer-aided-classification (CAD/CAC) processing string has been developed. This robust automated processing string involves the fusion of the outputs of unique mine classification algorithms. The overall CAD/CAC processing string consists of pre-processing, adaptive clutter filtering (ACF), normalization, detection, feature extraction, optimal subset feature selection, feature orthogonalization, classification and fusion processing blocks. The range-dimension ACF is matched both to average highlight and shadow information, while also adaptively suppressing background clutter. For each detected object, features are extracted and processed through an orthogonalization transformation, enabling an efficient application of the optimal log-likelihood-ratio-test (LLRT) classification rule, in the orthogonal feature space domain. The classified objects of 4 distinct processing strings are fused using the classification confidence values as features and “M-out-of-N”, or LLRT-based fusion rules. The utility of the overall processing strings and their fusion was demonstrated with new shallow water high-resolution sonar imagery data. The processing string detection and classification parameters were tuned and the string classification performance was optimized, by appropriately selecting a subset of the original feature set. Two significant improvements were made to the CAD/CAC processing string by employing sub-image adaptive clutter filtering (SACF) and utilizing a repeated application of the subset feature selection/feature orthogonalization/LLRT classification blocks. It was shown that LLRT-based fusion of the CAD/CAC processing strings outperforms the “M-out-of-N” algorithms and results in up to a seven-fold false alarm rate reduction, compared to the best single CAD/CAC processing string results, while maintaining a high correct mine classification probability. Alternately, the fusion of the processing strings enabled correct classification of almost all mine targets, while simultaneously maintaining a very low false alarm rate. A novel investigation was also presented that illustrates and provides insights on the increased performance gains provided by utilizing LLRT-based fusion of all the different combinations of 2, 3 or 4 distinct processing strings.

Proceedings Article | 21 September 2004 Paper

Robust score-based feature vectors for algorithm fusion

Gerald Dobeck

Proceedings Volume 5415, (2004) https://doi.org/10.1117/12.542731

KEYWORDS: Detection and tracking algorithms, Data fusion, Target detection, Matrices, Warfare, Sensors, Naval mines, Telecommunications, Receivers, Chemical elements

Read Abstract +

Proceedings Article | 11 September 2003 Paper

False alarm reduction by the And-ing of multiple multivariate Gaussian classifiers

Gerald Dobeck, J. Tory Cobb

Proceedings Volume 5089, (2003) https://doi.org/10.1117/12.487782

KEYWORDS: Feature selection, Naval mines, Matrices, Detection and tracking algorithms, Mining, Land mines, Algorithm development, Automatic target recognition, Neural networks, Image classification

Read Abstract +

Proceedings Article | 11 September 2003 Paper

Recent processing string and fusion algorithm improvements for automated sea mine classification in shallow water

Gerald Dobeck, Tom Aridgides, Manuel Fernandez

Proceedings Volume 5089, (2003) https://doi.org/10.1117/12.487775

KEYWORDS: Detection and tracking algorithms, Feature extraction, Data fusion, Algorithm development, Mining, Image processing, Naval mines, Image fusion, Feature selection, Image classification

Read Abstract +

Proceedings Article | 11 September 2003 Paper

Application of fusion algorithms for computer-aided detection and classification of bottom mines to shallow water test data from the battle space preparation autonomous underwater vehicle (BPAUV)

Gerald Dobeck, William Zurawski, Charles Ciany

Proceedings Volume 5089, (2003) https://doi.org/10.1117/12.487581

KEYWORDS: Detection and tracking algorithms, Image fusion, Data fusion, Image classification, Image processing, Image segmentation, Image processing algorithms and systems, Land mines, Algorithm development, Classification systems

Read Abstract +

Proceedings Article | 11 September 2003 Paper

Comparison of support vector machines and multilayer perceptron networks in building mine classification models

Martin Bello, Gerald Dobeck

Proceedings Volume 5089, (2003) https://doi.org/10.1117/12.487175

KEYWORDS: Detection and tracking algorithms, Mining, Feature selection, Land mines, Algorithm development, Neural networks, Data modeling, Image segmentation, Image filtering, Performance modeling

Read Abstract +

Proceedings Article | 13 August 2002 Paper

Fusion of multiple quadratic penalty function support vector machines (QPFSVM) for automated sea mine detection and classification

J. Tory Cobb, Gerald Dobeck

Proceedings Volume 4742, (2002) https://doi.org/10.1117/12.479112

KEYWORDS: Detection and tracking algorithms, Mining, Naval mines, Algorithm development, Land mines, Classification systems, Image filtering, Sensors, Surgery, Image processing

Read Abstract +

Proceedings Article | 13 August 2002 Paper

Application of fusion algorithims for computer-aided detection and classification of bottom mines to shallow-water test data

William Zurawski, Gerald Dobeck, Charles Ciany

Proceedings Volume 4742, (2002) https://doi.org/10.1117/12.479113

KEYWORDS: Detection and tracking algorithms, Image classification, Image fusion, Data fusion, Target detection, Image processing, Image segmentation, Land mines, Classification systems, Image processing algorithms and systems

Read Abstract +

Proceedings Article | 13 August 2002 Paper

Adaptive underwater target classification using a K-NN-based multi-aspect decision feedback unit

Khashayar Khorasani, Mahmood Azimi-Sadjadi, Arta Jamshidi, Gerald Dobeck

Proceedings Volume 4742, (2002) https://doi.org/10.1117/12.479117

KEYWORDS: Classification systems, Image classification, Lithium, Neural networks, Distance measurement, Feature extraction, Telecommunications, Environmental sensing, Acoustics, Autoregressive models

Read Abstract +

Proceedings Article | 13 August 2002 Paper

Fusion approach investigation for sea mine classification in very shallow water

Gerald Dobeck, Tom Aridgides, Manuel Fernandez

Proceedings Volume 4742, (2002) https://doi.org/10.1117/12.479114

KEYWORDS: Detection and tracking algorithms, Image processing, Data fusion, Feature extraction, Algorithm development, Mining, Image fusion, Naval mines, Land mines, Digital filtering

Read Abstract +

Proceedings Article | 18 October 2001 Paper

Side-scan sonar imagery fusion for sea mine detection and classification in very shallow water

Manuel Fernandez, Tom Aridgides, Gerald Dobeck

Proceedings Volume 4394, (2001) https://doi.org/10.1117/12.445440

KEYWORDS: Detection and tracking algorithms, Image fusion, Data fusion, Feature extraction, Algorithm development, Image processing, Image classification, Naval mines, Digital filtering, Feature selection

Read Abstract +

Proceedings Article | 18 October 2001 Paper

Comparison of confidence level of different classification paradigms for underwater target discrimination

Arta Jamshidi, Donghui Li, Mahmood Azimi-Sadjadi, Gerald Dobeck

Proceedings Volume 4394, (2001) https://doi.org/10.1117/12.445443

KEYWORDS: Neural networks, Distance measurement, Error analysis, Neurons, Acoustics, Feature extraction, Wavelets, Autoregressive models, Statistical analysis, Classification systems

Read Abstract +

Proceedings Article | 18 October 2001 Paper

On the power of algorithm fusion

Gerald Dobeck

Proceedings Volume 4394, (2001) https://doi.org/10.1117/12.445437

KEYWORDS: Detection and tracking algorithms, Algorithm development, Digital image correlation, Sensors, Sensor fusion, Naval mines, Data fusion, Target detection, Feature extraction, Image fusion

Read Abstract +

Proceedings Article | 18 October 2001 Paper

Adaptive underwater target classification with multi-aspect decision feedback

Mahmood Azimi-Sadjadi, Arta Jamshidi, Gerald Dobeck

Proceedings Volume 4394, (2001) https://doi.org/10.1117/12.445444

KEYWORDS: Classification systems, Neural networks, Environmental sensing, Feature extraction, Autoregressive models, Distance measurement, Acoustics, Sensors, Wavelets, Brain mapping

Read Abstract +

Proceedings Article | 22 August 2000 Paper

Algorithm fusion for the detection and classification of sea mines in the very shallow water region using side-scan sonar imagery

Gerald Dobeck

Proceedings Volume 4038, (2000) https://doi.org/10.1117/12.396262

KEYWORDS: Detection and tracking algorithms, Palladium, Digital image correlation, Image fusion, Naval mines, Algorithm development, Image classification, Image filtering, Mining, Target detection

Read Abstract +

Proceedings Article | 22 August 2000 Paper

Comparison of different neural network classification paradigms for underwater target discrimination

Gerald Dobeck, Donghui Li, Mahmood Azimi-Sadjadi

Proceedings Volume 4038, (2000) https://doi.org/10.1117/12.396261

KEYWORDS: Neural networks, Feature extraction, Classification systems, Acoustics, Neurons, Data modeling, Wavelets, Autoregressive models, Image classification, Environmental sensing

Read Abstract +

Proceedings Article | 22 August 2000 Paper

Comparison of effects of sonar bandwidth for underwater target classification

Mahmood Azimi-Sadjadi, De Yao, Donghui Li, Arta Jamshidi, Gerald Dobeck

Proceedings Volume 4038, (2000) https://doi.org/10.1117/12.396257

KEYWORDS: Classification systems, Data fusion, Neural networks, Error analysis, Wavelets, Mining, Monte Carlo methods, Feature extraction, Scattering, Signal processing

Read Abstract +

Proceedings Article | 22 August 2000 Paper

Fusion of sea mine detection and classification processing strings for sonar imagery

Tom Aridgides, Manuel Fernandez, Gerald Dobeck

Proceedings Volume 4038, (2000) https://doi.org/10.1117/12.396266

KEYWORDS: Detection and tracking algorithms, Algorithm development, Image processing, Image fusion, Data fusion, Feature extraction, Image classification, Naval mines, Digital filtering, Image filtering

Read Abstract +

Proceedings Article | 22 August 2000 Paper

Underwater target classification in changing environments using adaptive feature mapping schemes

Gerald Dobeck, De Yao, Donghui Li, Mahmood Azimi-Sadjadi

Proceedings Volume 4038, (2000) https://doi.org/10.1117/12.396259

KEYWORDS: Classification systems, Neural networks, Environmental sensing, Feature extraction, Wavelets, Sensors, Target detection, Acoustics, Signal processing, Autoregressive models

Read Abstract +

Proceedings Article | 2 August 1999 Paper

Adaptive underwater target classification system

Gerald Dobeck, De Yao, Mahmood Azimi-Sadjadi

Proceedings Volume 3710, (1999) https://doi.org/10.1117/12.357092

KEYWORDS: Neural networks, Classification systems, Feature extraction, Wavelets, Environmental sensing, Acoustics, Electronic filtering, Signal processing, Sensors, Submerged target modeling

Read Abstract +

Proceedings Article | 2 August 1999 Paper

Adaptive clutter suppression, sea mine detection/classification, and fusion processing string for sonar imagery

Manuel Fernandez, Tom Aridgides, Gerald Dobeck

Proceedings Volume 3710, (1999) https://doi.org/10.1117/12.357084

KEYWORDS: Image processing, Sensors, Image fusion, Target detection, Digital filtering, Land mines, Data fusion, Feature extraction, Filtering (signal processing), Naval mines

Read Abstract +

Proceedings Article | 2 August 1999 Paper

Feature extraction and fusion of wideband backscattered signals

Nathan Intrator, Gerald Dobeck, Quyen Huynh

Proceedings Volume 3710, (1999) https://doi.org/10.1117/12.357090

KEYWORDS: Wavelets, Feature extraction, Error analysis, Backscatter, Fermium, Frequency modulation, Distance measurement, Data modeling, Signal to noise ratio, Brain

Read Abstract +

Proceedings Article | 2 August 1999 Paper

Fusing sonar images for mine detection and classification

Gerald Dobeck

Proceedings Volume 3710, (1999) https://doi.org/10.1117/12.357082

KEYWORDS: Image fusion, Image classification, Image processing, Image filtering, Algorithm development, Mining, Databases, Detection and tracking algorithms, Land mines, Promethium

Read Abstract +

Proceedings Article | 4 September 1998 Paper

Underwater target classification using multi-aspect fusion and neural networks

Qiang Huang, Mahmood Azimi-Sadjadi, Gerald Dobeck

Proceedings Volume 3392, (1998) https://doi.org/10.1117/12.324142

KEYWORDS: Neural networks, Classification systems, Acoustics, Signal to noise ratio, Signal processing, Wavelets, Feature extraction, Data fusion, Target detection, Mining

Read Abstract +

Proceedings Article | 4 September 1998 Paper

Image enhancement for pattern recognition

Quyen Huynh, Gerald Dobeck, Nathan Intrator, Nicola Neretti

Proceedings Volume 3392, (1998) https://doi.org/10.1117/12.324202

KEYWORDS: Denoising, Wavelets, Image filtering, Gaussian filters, Image enhancement, Sensors, Detection and tracking algorithms, Pattern recognition, Warfare, Data centers

Read Abstract +

Proceedings Article | 4 September 1998 Paper

Adaptive clutter suppression and fusion processing string for sea mine detection and classification in sonar imagery

Manuel Fernandez, Gerald Dobeck, Tom Aridgides

Proceedings Volume 3392, (1998) https://doi.org/10.1117/12.324196

KEYWORDS: Image fusion, Digital filtering, Image processing, Target detection, Mining, Image classification, Sensors, Land mines, Data fusion, Filtering (signal processing)

Read Abstract +

An advanced, automatic, adaptive clutter suppression, pre- and post-detection level fusion, sea mine detection and classification processing string has been developed and tested with sonar imagery data. The overall string includes preprocessing, adaptive clutter filtering (ACF), normalization, detection, feature extraction and classification processing blocks. The ACF is a multi- dimensional adaptive linear FIR filter, optimal in the Least Squares sense, and is applied to low resolution data. It performs simultaneous background clutter suppression and preservation of an average peak target signature (normalized shape of mine highlights -- computed a priori using training set data). After data alignment, using a 3-dimensional ACF enables simultaneous multiple frequency data fusion and clutter suppression in the composite frequency-range- crossrange domain. Following 2-d normalization, the detection consists of thresholding, clustering of exceedances and limiting the number of detections. Finally, features are extracted from high resolution data and a orthogonalization transformation is applied to the features, enabling an efficient application of the optimal log-likelihood-ratio-test (LLRT) classification rule. Various fusion approaches were investigated including pre-detection fusion using the 3-d ACF string, post-detection fusion of the 2d ACF strings and an overall fusion of the two previous strings. The utility of the various processing strings was demonstrated with two new shallow water high resolution sonar imagery data sets. The overall ACF, pre- and post-detection level fusion, feature orthogonalization, LLRT-based classification processing string provided mine classification capability and false alarm rate performance exceeding the one of an expert sonar operator. A wide-sense stationary covariance model was utilized in the ACF algorithm design, significantly reducing the algorithm implementation complexity, thus enabling an easy implementation of the overall processing string in real-time.