15 Matching Results

Search Results

Advanced search parameters have been applied.

Intelligent Software Tools for Advanced Computing

Description: Feature extraction and evaluation are two procedures common to the development of any pattern recognition application. These features are the primary pieces of information which are used to train the pattern recognition tool, whether that tool is a neural network, a fuzzy logic rulebase, or a genetic algorithm. Careful selection of the features to be used by the pattern recognition tool can significantly streamline the overall development and training of the solution for the pattern recognition application. This report summarizes the development of an integrated, computer-based software package called the Feature Extraction Toolbox (FET), which can be used for the development and deployment of solutions to generic pattern recognition problems. This toolbox integrates a number of software techniques for signal processing, feature extraction and evaluation, and pattern recognition, all under a single, user-friendly development environment. The toolbox has been developed to run on a laptop computer, so that it may be taken to a site and used to develop pattern recognition applications in the field. A prototype version of this toolbox has been completed and is currently being used for applications development on several projects in support of the Department of Energy.
Date: April 3, 2001
Creator: Baumgart, C.W.
Partner: UNT Libraries Government Documents Department

Taxonomy of gamma ray burster data using a self-organizing neural network

Description: This paper summarizes the results of a signal taxonomy study of gamma ray burst (GRB) data acquired with sensors on-board the Pioneer-Venus Orbiter (PVO) spacecraft. GRB events produce large fluxes of gamma rays with durations of seconds to minutes and have been observed since the early 1970`s. The true nature of GRBs is still unknown and several competing theories exist. A fundamental point of contention among such theories is whether or not different types of GRB exist. If different types of GRBs are discovered in the existing PVO data base, the differences may correlate with their position or source characteristics. Hence, the goal of this project was to use artificial neural networks to perform signal taxonomy on the GRB data base to determine if unique classes or types of GRBs exist. A total of 26 signal features were identified, some of which can be associated directly with some characteristic of the GRB, such as duration, peak count rate, and gamma ray spectrum hardness. Additional features that were selected included the number of zero crossings in the wavelet transform and the fractal dimension of each signal. A self organizing neural network was used with the signal features to search for correlations among the signals contained in the database. The results of this analysis revealed an intrinsic dimensionality of 2 or 3 in the database. That is, it appears as though 2 or 3 distinct types of GRB may exist. In particular, two of the classes contain roughly 90% of the signals in the database of GRB signals we had to work with. These two classes are similar in characteristics but are still sufficiently distinct from one another to form separate categories. The third class of GRB is definitely distinct from the first two.
Date: December 31, 1993
Creator: Baumgart, C. W.
Partner: UNT Libraries Government Documents Department

An automated target recognition technique for image segmentation and scene analysis

Description: Automated target recognition software has been designed to perform image segmentation and scene analysis. Specifically, this software was developed as a package for the Army`s Minefield and Reconnaissance and Detector (MIRADOR) program. MIRADOR is an on/off road, remote control, multi-sensor system designed to detect buried and surface-emplaced metallic and non-metallic anti-tank mines. The basic requirements for this ATR software were: (1) an ability to separate target objects from the background in low S/N conditions; (2) an ability to handle a relatively high dynamic range in imaging light levels; (3) the ability to compensate for or remove light source effects such as shadows; and (4) the ability to identify target objects as mines. The image segmentation and target evaluation was performed utilizing an integrated and parallel processing approach. Three basic techniques (texture analysis, edge enhancement, and contrast enhancement) were used collectively to extract all potential mine target shapes from the basic image. Target evaluation was then performed using a combination of size, geometrical, and fractal characteristics which resulted in a calculated probability for each target shape. Overall results with this algorithm were quite good, though there is a trade-off between detection confidence and the number of false alarms. This technology also has applications in the areas of hazardous waste site remediation, archaeology, and law enforcement.
Date: February 1, 1994
Creator: Baumgart, C. W. & Ciarcia, C. A.
Partner: UNT Libraries Government Documents Department

Detecting errors and anomalies in computerized materials control and accountability databases

Description: The Automated MC and A Database Assessment project is aimed at improving anomaly and error detection in materials control and accountability (MC and A) databases and increasing confidence in the data that they contain. Anomalous data resulting in poor categorization of nuclear material inventories greatly reduces the value of the database information to users. Therefore it is essential that MC and A data be assessed periodically for anomalies or errors. Anomaly detection can identify errors in databases and thus provide assurance of the integrity of data. An expert system has been developed at Los Alamos National Laboratory that examines these large databases for anomalous or erroneous data. For several years, MC and A subject matter experts at Los Alamos have been using this automated system to examine the large amounts of accountability data that the Los Alamos Plutonium Facility generates. These data are collected and managed by the Material Accountability and Safeguards System, a near-real-time computerized nuclear material accountability and safeguards system. This year they have expanded the user base, customizing the anomaly detector for the varying requirements of different groups of users. This paper describes the progress in customizing the expert systems to the needs of the users of the data and reports on their results.
Date: December 31, 1998
Creator: Whiteson, R.; Hench, K.; Yarbro, T. & Baumgart, C.
Partner: UNT Libraries Government Documents Department

Imaging System for the Automated Determination of Microscopical Properties in Hardened Portland Concrete

Description: During this CRADA, Honeywell FM and T and MoDOT personnel designed a unique scanning system (including both hardware and software) that can be used to perform an automated scan and evaluation of a concrete sample. The specific goals of the CRADA were: (1) Develop a combined system integration, image acquisition, and image analysis approach to mimic the manual scanning and evaluation process. Produce a prototype system which can: (a) automate the scanning process to improve its speed and efficiency; (b) reduce operator fatigue; and (c) improve the consistency of the evaluation process. (2) Capture and preserve the baseline knowledge used by the MoDOT experts in performing the evaluation process. At the present time, the evaluation expertise resides in two MoDOT personnel. Automation of the evaluation process will allow that knowledge to be captured, preserved, and used for training purposes. (3) Develop an approach for the image analysis which is flexible and extensible in order to accommodate the inevitable pathologies that arise in the evaluation process. Such pathologies include features such as cracks and fissures, voids filled with paste or debris, and multiple, overlapping voids. FM and T personnel used image processing, pattern recognition, and system integration skills developed for other Department of Energy applications to develop and test a prototype of an automated scanning system for concrete evaluation. MoDOT personnel provided all the basic hardware (microscope, camera, computer-controlled stage, etc.) for the prototype, supported FM and T in the acquisition of image data for software development, and provided their critical expert knowledge of the process of concrete evaluation. This combination of expertise was vital to the successful development of the prototype system.
Date: March 8, 2000
Creator: Baumgart, C.W.; Cave, S.P. & Linder, K.E.
Partner: UNT Libraries Government Documents Department

A Feature Extraction Toolbox for Pattern Recognition Application

Description: Feature extraction and evaluation are procedures common to the development of all pattern recognition application. These features are the primary pieces of information used to train the pattern recognition engine, whether that engine is a neural network, a fuzzy logic rulebase, or a genetic algorithm. Careful selection of the features to be used by the pattern recognition engine can significantly streamline the overall development and training of the solution for the pattern recognition application. Presently, AlliedSignal Federal Manufacturing & Technologies (FM&T) is developing an integrated, computer-based software package, called the Feature Extraction Toolbox. This package will be used for developing and deploying solutions to generic pattern recognition problems. The toolbox integrates a variety of software techniques for signal processing, feature extraction and evaluation, and pattern recognition, under a single, user-friendly developmental environment. While a feature extraction toolbox can help in the selection process, it is the user that ultimately must make all decisions. A prototype version of this toolbox has been developed and currently is being used for applications development on several projects in support of the Department of Energy. The toolbox has been developed to run on a laptop computer so that it can be taken to a site and used in the field.
Date: November 23, 1998
Creator: Baumgart, C.W.; Linder, K.E. & Nelson, L.K.
Partner: UNT Libraries Government Documents Department

An Automated Image Processing System for Concrete Evaluation

Description: AlliedSignal Federal Manufacturing & Technologies (FM&T) was asked to perform a proof-of-concept study for the Missouri Highway and Transportation Department (MHTD), Research Division, in June 1997. The goal of this proof-of-concept study was to ascertain if automated scanning and imaging techniques might be applied effectively to the problem of concrete evaluation. In the current evaluation process, a concrete sample core is manually scanned under a microscope. Voids (or air spaces) within the concrete are then detected visually by a human operator by incrementing the sample under the cross-hairs of a microscope and by counting the number of "pixels" which fall within a void. Automation of the scanning and image analysis processes is desired to improve the speed of the scanning process, to improve evaluation consistency, and to reduce operator fatigue. An initial, proof-of-concept image analysis approach was successfully developed and demonstrated using acquired black and white imagery of concrete samples. In this paper, the automated scanning and image capture system currently under development will be described and the image processing approach developed for the proof-of-concept study will be demonstrated. A development update and plans for future enhancements are also presented.
Date: November 23, 1998
Creator: Baumgart, C.W.; Cave, S.P. & Linder, K.E.
Partner: UNT Libraries Government Documents Department

Performance Enhancement of the Automated Concrete Evaluation System (ACES)

Description: The objective of this proposed research is to improve and expand the detection and analysis capabilities of the automated, concrete evaluation (ACE) system. MoDOT and Honeywell jointly developed this system. The focus of this proposed research will be on the following: Coordination of concrete imaging efforts with other states, Validation and testing of the ACE system on a broad range of concrete samples, and Identification and development of software and hardware enhancements. These enhancements will meet the needs of diverse users in the field of concrete materials, construction, and research.
Date: February 14, 2002
Creator: Baumgart,C.W.; Cave,S.P. & Linder,K.E.
Partner: UNT Libraries Government Documents Department

A vehicle mounted multi-sensor array for waste site characterization

Description: Personnel at AlliedSignal Aerospace, Kirtland Operations (formerly EG&G Energy Measurements, Kirtland Operations) and EG&G Energy Measurements, Los Alamos Operations, have successfully developed and demonstrated a number of technologies which can be applied to the environmental remediation and waste management problem. These applications have included the development of self-contained and towed remote sensing platforms and advanced signal analysis techniques for the detection and characterization of subsurface features. This presentation will provide a brief overview of applications that have been and are currently being fielded by both AlliedSignal and EG&G Energy Measurements personnel and will describe some of the ways that such technologies can and are being used for the detection and characterization of hazardous waste sites.
Date: February 1, 1995
Creator: Baumgart, C. W.; Ciarcia, C. A. & Tunnell, T. W.
Partner: UNT Libraries Government Documents Department

An anomaly detector applied to a materials control and accounting system

Description: Large amounts of safeguards data are automatically gathered and stored by monitoring instruments used in nuclear chemical processing plants, nuclear material storage facilities, and nuclear fuel fabrication facilities. An integrated safeguards approach requires the ability to identify anomalous activities or states in these data. Anomalies in the data could be indications of error, theft, or diversion of material. The large volume of the data makes analysis and evaluation by human experts very tedious, and the complex and diverse nature of the data makes these tasks difficult to automate. This paper describes the early work in the development of analysis tools to automate the anomaly detection process. Using data from accounting databases, the authors are modeling the normal behavior of processes. From these models they hope to be able to identify activities or data that deviate from that norm. Such tools would be used to reveal trends, identify errors, and recognize unusual data. Thus the expert`s attention can be focused directly on significant phenomena.
Date: August 1, 1994
Creator: Whiteson, R.; Kelso, F.; Baumgart, C. & Tunnell, T. W.
Partner: UNT Libraries Government Documents Department

An integrated, subsurface characterization system for real-time, in-situ field analysis

Description: This paper describes current efforts at AlliedSignal Federal Manufacturing and Technologies (FM and T) to develop and field an in-situ, data analysis platform to acquire, process, and display site survey data in near real-time. In past years, FM and T has performed a number of site survey tasks. Each of these surveys was unique in application as well as in the type of data processing and analysis that was required to extract and visualize useful site characterization information. However, common to each of these surveys were the following specific computational and operational requirements: (1) a capability to acquire, process, and visualize the site survey data in the field; (2) a capability to perform all processing in a timely fashion (ideally real-time); and (3) a technique for correlating (or fusing) data streams from multiple sensors. Two more general, but no less important, requirements include system architecture modularity and positioning capability. Potential applications include: survey, evaluation, and remediation of numerous Department of Defense and Department of Energy waste sites; real-time detection and characterization of unexploded ordnance and landmines; survey, evaluation, and remediation of industrial waste sites; location of underground utility lines; and providing law enforcement agencies with real-time surveys of crime scenes. The paper describes an integrated data acquisition, processing, and visualization platform that is capable of performing in-situ data processing, interpretation, and visualization in real-time.
Date: February 1, 1996
Creator: Baumgart, C.W.; Creager, J.; Mathes, J.; Pounds, T.; VanDeusen, A. & Warthen, B.
Partner: UNT Libraries Government Documents Department

Automated detection of Karnal bunt teliospores

Description: Karnal bunt is a fungal disease which infects wheat and, when present in wheat crops, yields it unsatisfactory for human consumption. Due to the fact that Karnal bunt (KB) is difficult to detect in the field, samples are taken to laboratories where technicians use microscopes and methodically search for KB teliospores. AlliedSignal Federal Manufacturing and Technologies (FM and T), working with the Kansas Department of Agriculture, created a system which utilizes pattern recognition, feature extraction, and neural networks to prototype an automated detection system for identifying KB teliospores. System hardware consists of a biological compound microscope, motorized stage, CCD camera, frame grabber, and a PC. Integration of the system hardware with custom software comprises the machine vision system. Fundamental processing steps involve capturing an image from the slide, while concurrently processing the previous image. Features extracted from the acquired imagery are then processed by a neural network classifier which has been trained to recognize spore-like objects. Images with spore-like objects are reviewed by trained technicians. Benefits of this system include: (1) reduction of the overall cycle-time; (2) utilization of technicians for intelligent decision making (vs. manual searching); (3) a regulatory standard which is quantifiable and repeatable; (4) guaranteed 100% coverage of the cover slip; and (5) significantly enhanced detection accuracy.
Date: February 1, 1998
Creator: Linder, K. D.; Baumgart, C.; Creager, J.; Heinen, B.; Troupe, T.; Meyer, D. et al.
Partner: UNT Libraries Government Documents Department

Integration of video and radiation analysis data

Description: For the past several years, the integration of containment and surveillance (C/S) with nondestructive assay (NDA) sensors for monitoring the movement of nuclear material has focused on the hardware and communications protocols in the transmission network. Little progress has been made in methods to utilize the combined C/S and NDA data for safeguards and to reduce the inspector time spent in nuclear facilities. One of the fundamental problems in the integration of the combined data is that the two methods operate in different dimensions. The C/S video data is spatial in nature; whereas, the NDA sensors provide radiation levels versus time data. The authors have introduced a new method to integrate spatial (digital video) with time (radiation monitoring) information. This technology is based on pattern recognition by neural networks, provides significant capability to analyze complex data, and has the ability to learn and adapt to changing situations. This technique has the potential of significantly reducing the frequency of inspection visits to key facilities without a loss of safeguards effectiveness.
Date: December 31, 1995
Creator: Menlove, H. O.; Howell, J. A.; Rodriguez, C. A.; Eccleston, G. W.; Beddingfield, D.; Smith, J. E. et al.
Partner: UNT Libraries Government Documents Department

A space-based classification system for RF transients

Description: The FORTE (Fast On-Orbit Recording of Transient Events) small satellite is scheduled for launch in mid 1995. The mission is to measure and classify VHF (30--300 MHz) electromagnetic pulses, primarily due to lightning, within a high noise environment dominated by continuous wave carriers such as TV and FM stations. The FORTE Event Classifier will use specialized hardware to implement signal processing and neural network algorithms that perform onboard classification of RF transients and carriers. Lightning events will also be characterized with optical data telemetered to the ground. A primary mission science goal is to develop a comprehensive understanding of the correlation between the optical flash and the VHF emissions from lightning. By combining FORTE measurements with ground measurements and/or active transmitters, other science issues can be addressed. Examples include the correlation of global precipitation rates with lightning flash rates and location, the effects of large scale structures within the ionosphere (such as traveling ionospheric disturbances and horizontal gradients in the total electron content) on the propagation of broad bandwidth RF signals, and various areas of lightning physics. Event classification is a key feature of the FORTE mission. Neural networks are promising candidates for this application. The authors describe the proposed FORTE Event Classifier flight system, which consists of a commercially available digital signal processing board and a custom board, and discuss work on signal processing and neural network algorithms.
Date: December 1, 1993
Creator: Moore, K. R.; Call, D.; Johnson, S.; Payne, T.; Ford, W.; Spencer, K. et al.
Partner: UNT Libraries Government Documents Department

Integration of video and radiation analysis data

Description: We have introduced a new method to integrate spatial (digital video) and time (radiation monitoring) information. This technology is based on pattern recognition by neural networks, provides significant capability to analyze complex data, and has the ability to learn and adapt to changing situations. This technique could significantly reduce the frequency of inspection visits to key facilities without a loss of safeguards effectiveness.
Date: August 1, 1994
Creator: Menlove, H. O.; Howell, J. A.; Rodriguez, C. A.; Eccleston, G. W.; Beddingfield, D.; Smith, J. E. et al.
Partner: UNT Libraries Government Documents Department