A new optoelectronic technology for the implementation of neural network architectures is being developed at Lincoln Laboratory. The new technology is based on a multiple-quantum- well (MQW) device called the monolith...
详细信息
ISBN:
(纸本)0819405787
A new optoelectronic technology for the implementation of neural network architectures is being developed at Lincoln Laboratory. The new technology is based on a multiple-quantum- well (MQW) device called the monolithic optoelectronic transistor (MOET). MOET is a true optical transistor;it enables the switching of one optical signal with a much weaker one. The terminal characteristics are ideal for implementing a neuron: abrupt or sigmoidal thresholds, saturated ON and OFF states, and high fan-out. The device will be initially demonstrated for implementation of early visual processingnetworks. The baseline network is the CORT-X model, a multiple spatial-scale, feedforward network for boundary segmentation of noisy binary images. MOET implementation is possible with slight modifications to the CORT-X architecture. Simulations of the hardware implemented network are carried out and compared with the performance of the original model. As a function of input image carrier-to-noise ratio (CNR), performance is evaluated with respect to deviations from ideal response along two dimensions: (1) contrast-ratio and (2) nonuniformity. Assuming ideal hardware response, the modified CORT-X architecture performs better than the original model. Moderate contrast does not significantly degrade network performance, while nonuniformities as small as 10% degrade performance even for high CNR.
A comparative study of the artificialneural computing and traditional approaches to imageprocessing is performed. The major goal is to determine the usefulness of artificialneural systems (ANSs) for such image proc...
详细信息
A comparative study of the artificialneural computing and traditional approaches to imageprocessing is performed. The major goal is to determine the usefulness of artificialneural systems (ANSs) for such imageprocessingapplications as histogramming and image encoding. The paradigm used was developed from a C programming language model of a perceptron ANS with consideration of backpropagation attributes. It is found that the ANS approach produces results similar to those of traditional techniques.< >
Describes a range of neural signal processing methods employed for B-Scan ultrasonic image enhancement and material identification. All approaches assume no a-priori knowledge of the environment. A Multi-Layered Perce...
详细信息
Describes a range of neural signal processing methods employed for B-Scan ultrasonic image enhancement and material identification. All approaches assume no a-priori knowledge of the environment. A Multi-Layered Perceptron (MLP) employing back propagation learning was used for all aspects of this research. The motivation for this work arises from a requirement to map and navigate within, hazardous environments. Ultrasonic transducers have advantages in such circumstances due to their mechanical robustness and low replacement cost.< >
The architecture of an intelligent imageprocessing system (IIMS) is analyzed, and the main problems related to its practical implementation are discussed. A recent trend in imageprocessing and analysis is borrowing ...
详细信息
The architecture of an intelligent imageprocessing system (IIMS) is analyzed, and the main problems related to its practical implementation are discussed. A recent trend in imageprocessing and analysis is borrowing from the fields of artificial intelligence, pattern recognition, and neuralnetworks in order to make it possible to automatically extract various kinds of information from static and dynamic images. The fundamental paradigms relative to the intelligent processing of images are analyzed, particular attention is devoted: to the data fusion function which initiate the process and gives information on the actual situation with the associated uncertainties; to the hypotheses function, which takes into consideration the uncertainties in the correlated data and creates the best explanations of the fused data; to the option function, which creates various response alternatives for each hypothesis simulating the resulting effects; and to the response function, which realizes the planned actions so producing new stimuli and initializing other processing cycles.< >
作者：
WEBSTER, WPThe Author The Author is the technology manager for the Intercept Weapons Department at the Naval Weapons Center
China Lake Calif. Dr. Webster received his B.S. M.S. and Ph.D. degrees in electrical engineering at the University of Arizona Tucson Arizona in 1964 1966 and 1971 respectively. While at the University of Arizona his studies focused on electromagnetic field propagation and scattering and laser radar system design. Since graduation he has worked at the Naval Weapons Center in a variety of technology areas including laser radar system design infrared system design proximity fuze system design hybrid and monolithic microelectronic circuit design and fabrication technology airborne fire control sensor technology and anti-air and anti-surface missile guidance and control technology. Currently he is responsible for identifying selecting and directing the development of advanced guidance sensor signal processing and aerodynamic control system technologies for application to U.S. Navy tactical electro-optical and radio frequency air intercept missiles. Dr. Webster is a member of IEEE Sigma Xi and Tau Beta Pi.
Complex processing algorithms associated with requirements for real-time target detection, acquisition and recognition have far outdistanced our ability to package the necessary processing power into real-time weapon-...
详细信息
Complex processing algorithms associated with requirements for real-time target detection, acquisition and recognition have far outdistanced our ability to package the necessary processing power into real-time weapon-sized hardware. Weapon developers need new, radically different, approaches to solve these difficult weapon signal processing problems. artificialneuralnetworks (ANN) is one of the new, radically different, signal processing approaches that is currently receiving much attention for weapon applications. The human brain is a powerful image and pattern recognition processor whose basic processing element is the neuron. Synapses are the weighted interconnections between neurons that permit learning and communication between the neurons. artificialneural network technology attempts to mathematically and/or electrically model neurons and synapses and then interconnect these models in architectures suitable for signal processing tasks. ANN technology is particularly applicable to pattern recognition, speech recognition, machine vision, robotics, and optimization signal processing tasks. Specific military applications include missile seekers, missile fuzing, sonar target discrimination, automatic target recognition, and autopilots. Two unique characteristics of ANN processors are that they are non-linear processors and that they are trained, not programmed, to accomplish processing tasks in a manner analogous to the way the human brain learns. Learning is achieved by modifying the synaptic weights of each artificial neuron until the final desired system processing response is achieved. Several years ago the Naval Weapons Center (NavWpnCen) identified ANN technology as a high risk-high payoff approach to missile signal processing requirements and initiated several research and development efforts in this area. One result of this work is the new 80170NW ANN analog VLSI chip produced by Intel Inc. Another is the current Missileborne Integrated neural Network De
Hype has played a significant role in the history of artificialneural network research. From the initial recognition of the potential for networks to approximate non-linear functions some quite exaggerated claims hav...
详细信息
Hype has played a significant role in the history of artificialneural network research. From the initial recognition of the potential for networks to approximate non-linear functions some quite exaggerated claims have been an unwelcome accompaniment to research developments. Results from both simulation studies and real process applications indicated that even simple network architectures appeared to possess powerful non-linear process modelling capabilities that provided easy and quick practical solutions to complex problems Optimism grew and the pragmatic solution was borne. In contrast, the structure of a neural network based model was also being considered generic in the sense that little prior knowledge of the process was required. Unlike ARMAX and NARMAX approaches the methodology has been attributed the potential of accurately describing the behaviour of extremely complex systems. Claims that the technique offered a panacea to all modelling problems surfaced. This paper considers a number of case studies and examines the justification for these claims.
An application of neocognitron to the formation of a closed boundary contour of the left ventricle (LV) from line segments in echocardiographic images is presented. In echocardiographic imageprocessing, the extractio...
详细信息
ISBN:
(纸本)0819405787
An application of neocognitron to the formation of a closed boundary contour of the left ventricle (LV) from line segments in echocardiographic images is presented. In echocardiographic imageprocessing, the extraction of a closed boundary contour is a very difficult but important task for both automatic diagnosing and monitoring of heart functions. The two major problems are the extraction of complex contour edge pattern structures from noisy edge streaks and the reconstruction of missing edges. The first problem can be cast as a feature extraction problem, and the second one can be solved by the top-down model-based approach using the existing contextual information. The neocognitron model which has a hierarchical structure and forward and backward paths provides solutions to these two problems. The recognition of an input pattern is carried out sequentially, layer by layer along the forward path. The forward signals route is later retraced by the backward signals. The noise edges in the input pattern, which are not retraced in the backward route, are eliminated by the gain control mechanism. The incomplete features, which appear in the backward route, but might not be extracted in the forward route, also can be extracted by lowering the threshold value. The missing edge in the initial pattern can be filled back once the feature corresponding to that edge has been reconstructed. This neocognitron network has three layers: the lower, the intermediate, and the highest, containing the templates for basic geometrical features, basic curve features, and whole boundary contour, respectively. The models of the left ventricle boundary are created semi-automatically from real echocardiographic images. A model is established for each class of images based on the shape of the boundary contour which is defined according to age, gender, disease, probe position, and the time of contraction or expansion. The curve features are extracted from the boundary contour model semi-autom
Architectures and detailed circuit designs of one analog trainable neural chip and one-digital systolic-processor chip are presented. The analog vector quantizer chip performs full search in a massively parallel fashi...
详细信息
Architectures and detailed circuit designs of one analog trainable neural chip and one-digital systolic-processor chip are presented. The analog vector quantizer chip performs full search in a massively parallel fashion with an expandable winner-take-all circuitry which can achieve a 10-b resolution. A high compression ratio of 33 is feasible in many image compression applications. Extensive design of a digital systolic-processor chip has been conducted. Circuit blocks, data communication, and microcodes are created to support either the ring-connected or the mesh-connected systolic array for the retrieving and learning phases of the neural network operation. The digital neural chip can also be configured to implement fuzzy logic systems.< >
A prototype system is being developed to demonstrate the use of neural network hardware to fuse multispectral imagery. This system consists of a neural network IC on a motherboard a circuit card assembly and a set of ...
详细信息
ISBN:
(纸本)0819403458
A prototype system is being developed to demonstrate the use of neural network hardware to fuse multispectral imagery. This system consists of a neural network IC on a motherboard a circuit card assembly and a set of software routines hosted by a PC-class computer. Research in support of this consists of neural network simulations fusing 4 to 7 bands of Landsat imagery and fusing (separately) multiple bands of synthetic imagery. The simulations results and a description of the prototype system are presented. 1.
暂无评论