In the weaving industry a variety of meshes are manufactured, using both metallic and nonmetallic materials, for a wide variety of filtration applications. At present the quality of the mesh is monitored at several st...
详细信息
In the weaving industry a variety of meshes are manufactured, using both metallic and nonmetallic materials, for a wide variety of filtration applications. At present the quality of the mesh is monitored at several stages of the manufacturing process to keep selected parameters within specified limits. This involves stopping the process approximately every half hour, manually assessing the mesh quality, and taking remedial action as necessary. There is a need to replace these manual checks with an automatic system that can continually monitor the mesh as it is being manufactured and, with appropriate actuators, control the loom on-line. The authors present initial results of research being conducted at Sheffield University to develop such a system for wire mesh. Early experiments proved that a contact measurement technique is inappropriate due to the severe vibrations and backlash nonlinearities present in most looms. The approach adopted here therefore, is to use imageprocessing techniques. The instrument is transputer based so that the imageprocessing can be performed on several processors to achieve the real-time performance necessary for on-line monitoring and control purposes.< >
Describes research into the specification, design and implementation of real-time synchronisation and control software for a Transputer-controlled, modular, multi-axis, independently driven, high-speed, packaging mach...
详细信息
Describes research into the specification, design and implementation of real-time synchronisation and control software for a Transputer-controlled, modular, multi-axis, independently driven, high-speed, packaging machine under normal and abnormal conditions. The research has already demonstrated the feasibility of using independent drives and software-based controllers for the synchronisation of both intermittent and continuous motions in applications at the boundary of existing technology.< >
The Warwick Pyramid Machine (WPM) is an M-SIMD (multiple-single instruction multiple data) heterogeneous pyramid architecture for real-time image understanding. SIMD (single instruction multiple data) machines are app...
详细信息
The Warwick Pyramid Machine (WPM) is an M-SIMD (multiple-single instruction multiple data) heterogeneous pyramid architecture for real-time image understanding. SIMD (single instruction multiple data) machines are appropriate for the early stages of imageprocessing. Subsequent numeric and symbolic processing makes poor use of an SIMD array, and is more suited to one or more conventional processors. Multiple-SIMD architectures divide an SIMD array into smaller sub-arrays, or clusters. Clusters may operate in synchrony, or autonomously. In the Warwick M-SIMD architecture each cluster consists of a small SIMD array, a controller, and a conventional microprocessor. The controller and the SIMD array may be considered as a single processor with a conventional ALU (arithmetic logic unit) and register structure, but with the addition of an array ALU with array registers. The conventional microprocessors form an MIMD (multiple instruction multiple data) array. Clusters may be configured to meet the real-time, or other, constraints of an application.< >
The generation of perspective views of 3D surfaces obtained from predetermined view points is a complex and computationaly expensive task requiring the storage, manipulation and transformation of large amounts of data...
详细信息
The generation of perspective views of 3D surfaces obtained from predetermined view points is a complex and computationaly expensive task requiring the storage, manipulation and transformation of large amounts of data. At present there are a number of conventional algorithms which take digital surface data and process it to produce such images. Current algorithms are inherently sequential and do not, in general, map easily onto parallel processor architectures. This paper describes an algorithm which exploits the data independent nature of the calculations for individual pixel columns of the image and describes how this algorithm can be efficiently mapped onto an array of processors. Using an inherently parallel algorithm perspective views can be produced many times faster and therefore can be used for new applications.< >
Residue number arithmetic is a technique for mapping an integer computation into several parallel and independent computations requiring much shorter wordlengths. In this paper the authors present an architecture for ...
详细信息
Residue number arithmetic is a technique for mapping an integer computation into several parallel and independent computations requiring much shorter wordlengths. In this paper the authors present an architecture for carrying out modulo arithmetic in a practically useful residue number system. An important feature of this is that it allows a single hardware design to be used for all moduli. They describe an application of this architecture for computing inner-products in convolutions, transforms and matrix-vector operations.< >
The fractal model is used as the basis of a texture analysis and synthesis technique for use on perceived visual texture in natural scenes of vegetation. Preliminary studies found that natural scenes depart from the f...
详细信息
The fractal model is used as the basis of a texture analysis and synthesis technique for use on perceived visual texture in natural scenes of vegetation. Preliminary studies found that natural scenes depart from the fractal model. The model is adapted to allow for deviation between perceived natural visual texture and true fractals. The texture generator is used to synthesize natural visual texture. A transputer array is used for fast parallel analysis of the perceived visual texture of natural scenes for robot vision.< >
The transputer has been described as a 'building block' for parallelprocessing systems. The IMS B420 (Vectram) has been developed to offer both scalar and vector processing in a single module. This module com...
详细信息
The transputer has been described as a 'building block' for parallelprocessing systems. The IMS B420 (Vectram) has been developed to offer both scalar and vector processing in a single module. This module combines the scalar floating/integer and the general purpose capabilities of the transputer with a high performance vector/signal processing co-processor. A set of software libraries are available which allow programs, written in high level languages executing on the transputer, to call vector routines which execute on the coprocessor. The author describes the Vectram and some applications.< >
The authors examine the basic characteristics of three different types of optical network in the context of important communications bottlenecks that may arise as the processing power of computers increases. They are ...
详细信息
The authors examine the basic characteristics of three different types of optical network in the context of important communications bottlenecks that may arise as the processing power of computers increases. They are more concerned with estimates of the ultimate performance of archetypal networks and computer architectures, e.g. optically connected parallel machines.< >
Most image analysis tasks are computationally intensive and this is a particularly important factor in applications where a primary consideration is to maximise operational speeds. Transputer arrays provide an infrast...
详细信息
Most image analysis tasks are computationally intensive and this is a particularly important factor in applications where a primary consideration is to maximise operational speeds. Transputer arrays provide an infrastructure for implementation of algorithms which can be matched quite closely to the inherent structural features embedded in typical task areas. The paper describes an approach to the specification of algorithms for image classification which embody a direct link with implementational considerations.< >
The York Probabilistic Automata Machine is a video rate 256 cube Automata Array. The system has been developed primarily in ASIC based components, with the bulk of processing performed through a set of overlapped pipe...
详细信息
The York Probabilistic Automata Machine is a video rate 256 cube Automata Array. The system has been developed primarily in ASIC based components, with the bulk of processing performed through a set of overlapped pipelines. Physical connectivity is limited to 128 nodes (in three dimensions), although higher 'virtual connectivities' are possible. Individual operations permitted at cellular level include binary and 'fuzzy' logic, plus simple weighted arithmetic. To complement the processor configuration, high bandwidth I/O and display units are currently under development, allowing arbitrary mappings of data I/O onto three dimensional surfaces within the unit. This paper introduces the York PAM and develops the application of virtual connectivity in automata like systems for massively parallel computation. The design and implementation of the individual processing element are presented, and the design of the three dimensional video re-sequencer discussed. Finally the use of probabilistic automata for the presentation of high level image data and facial recognition through neural techniques are presented.< >
暂无评论