Implementation issues of neural network formulations onto an available computer architectures are unique. In conventional signal and imageprocessing, for example, many of the frequently performed mathematical operati...
详细信息
ISBN:
(纸本)0819402931
Implementation issues of neural network formulations onto an available computer architectures are unique. In conventional signal and imageprocessing, for example, many of the frequently performed mathematical operations, like convolutions and transformations, are matrix-vector operations. Computational requirements of these algorithms are met by parallel hardware architectures tailored for efficient handling of vector data. However, many neural network systems require more than just making the computation parallel. Implementation of neural network architectures, such as those exemplified by Grossberg's Boundary Contour System, involve the solution of hundreds of coupled ordinary non-linear differential equations. A neurocomputer whose basic computational unit performs integration, rather the usual arithmetic operations, would be ideal for these systems. The training of the back propagation network, similarly, can be expressed as a problem of solving a system of stiff, coupled ordinary differential equations for the weights which are treated as continuous functions of time. The ability to efficiently solve differential equations on digital and parallel computers is therefore quite important for the implelementation of artificial neural networks. The central idea of a mapping, described in this paper, involves the replacement of differentiation operators and functions, in the given equation, respectively with the so-called differentiation matrices and function matrices. With the help of a so-called projection matrix, a differential equation is transformed into a rectangular vector-matrix equation which can then be solved on a systolic processor. The algorithm is computationally efficient and enables one to numerically compute, separately, the general solution for the homogeneous part and the particular solution for any specified forcing function.
This paper describes two techniques for automatic recognition of surface targets from an airborne platform using an imaging laser radar sensor. The first technique rotates a three-dimensional model of the target in re...
详细信息
This paper describes two techniques for automatic recognition of surface targets from an airborne platform using an imaging laser radar sensor. The first technique rotates a three-dimensional model of the target in re...
详细信息
This paper describes two techniques for automatic recognition of surface targets from an airborne platform using an imaging laser radar sensor. The first technique rotates a three-dimensional model of the target in real time to enable a generalized Hough transform to match the ladar image to the target's key discriminating features as a basis for target identification. The second technique uses a variation on minimum average correlation energy filters to perform robust target identification. Examples illustrating the application of these algorithms are presented, along with a description of a real-time implementation of the critical parts of the algorithms on a 40,000-processor systolic array based on the Geometric Arithmetic Parallel Processor (GAPPTM) chip.
Optimal algorithms for evaluating the amplitude-phase image of a spatially distributed statistically uneven and electrophysical nonuniform object in the presence of white Gaussian noise are synthesized under condition...
详细信息
Optimal algorithms for evaluating the amplitude-phase image of a spatially distributed statistically uneven and electrophysical nonuniform object in the presence of white Gaussian noise are synthesized under conditions of remote radio sounding. The optimal structures for space-time processing are determined. The advantages of the synthesized algorithms over typical radar stations with a synthesized aperture are pointed out.
An imageprocessing workstation called the vision and image engineering workstation, or vIEW-Station, which as developed in order to merge the state-of-the-art software environment of a UNIX workstation and the comput...
详细信息
An imageprocessing workstation called the vision and image engineering workstation, or vIEW-Station, which as developed in order to merge the state-of-the-art software environment of a UNIX workstation and the computing power of the fast image processor, is described. The software system of the vIEW-Station is constructed hierarchically in order to allow device independence in porting to various hardware configurations and to achieve flexibility in developing application systems. An image computing language called v-Sugar, which provides specific data types and a functional notation, allowing imageprocessingalgorithms to be expressed simply, is described. The image memory management mechanism, v-server, and the window system extensions, vIEW-Windows, are described. They are integrated to provide an independent basis to hardware configurations.< >
An algorithm for identifying contours on images is described. This algorithm is intended for use with digital display systems. It can accelerate computations by a factor of 100 in comparison with the standard algorith...
详细信息
An algorithm for identifying contours on images is described. This algorithm is intended for use with digital display systems. It can accelerate computations by a factor of 100 in comparison with the standard algorithms. With a slight modification, the algorithm can be used to rapidly identify other geometric structures on images. This algorithm can be used to solve practical problems in geology, agriculture, and other fields involving remote sensing of the Earth from space.
The concept of online processing is presented as an effective approach to overcome the data distribution overhead in parallel real-time implementation of low-level vision. Online processing refers to pipelining each i...
详细信息
The concept of online processing is presented as an effective approach to overcome the data distribution overhead in parallel real-time implementation of low-level vision. Online processing refers to pipelining each input item with the pipeline rate equal to the input arrival rate. Online algorithms are presented for four benchmark low level tasks, namely two-dimensional convolution, two-dimensional rank value filtering, connected component labeling, and Hough transform. A two-stage pipelined module that allows online implementations is described. This module is to function as a coprocessor to nodes in B-HIvE, a loosely coupled multiprocessor for integrated vision.< >
The study of Regular Iterative algorithms (RIAs) was introduced in a seminal paper by Karp, Miller, and Winograd in 1967. In more recent years, the study of systolic architectures has led to a renewed interest in this...
This paper describes a parallel, implementation of algorithms for automatic target identification using a CO2 laser radar sensor, which provides pixel-registered visual, infrared, relative range and Doppler (motion) i...
详细信息
暂无评论