Relaxation method attracts attention as an effective method in imageprocessing.and pattern recognition. From the first, the relaxation operation has been said to be capable of parallel processing. But previous hardwa...
详细信息
Relaxation method attracts attention as an effective method in imageprocessing.and pattern recognition. From the first, the relaxation operation has been said to be capable of parallel processing. But previous hardware implementations were specialized to each application. This paper presents a parallel architecture of a relaxation processor which can be applied to general use.
The paper presents a computer which employs massive parallelism. This massively parallel processing.computer (MPPC) manipulates arrays of 0s and 1s in contrast to conventional manipulation of bits. The proposed comput...
详细信息
The paper presents a computer which employs massive parallelism. This massively parallel processing.computer (MPPC) manipulates arrays of 0s and 1s in contrast to conventional manipulation of bits. The proposed computer will have a spatially parallel architecture that would enable it to recognize automatically in just milliseconds the details in satellite pictures. MPPC hardware is made up of 128 multiplied by 128 matrix of bit-serial processing.elements in an array unit (ARU) and its control - the array control unit (ACU). These are interfaced to a host controller computer (HCC). The instruction set can be divided into three basic classes viz. Boolean, simple near-neighbour, and recursive near neighbour.
This conference proceedings contains 14 papers, two of which are in abstract form only. The main subjects are local neighbourhood operators, display of digitalimage data, remote sensing, image thinning algorithms, re...
详细信息
ISBN:
(纸本)0869804731
This conference proceedings contains 14 papers, two of which are in abstract form only. The main subjects are local neighbourhood operators, display of digitalimage data, remote sensing, image thinning algorithms, realtime data compression, syntactic edge detection algorithms, parallel processing.image computers, computer vision systems, shape recognition, and characterisation of visual texture.
The problem of computing projection, of digitalimages is considered. algorithms are presented that are suitable for implementation in general-purpose imageprocessing.and image analysis pipeline architectures. Also, ...
详细信息
The problem of computing projection, of digitalimages is considered. algorithms are presented that are suitable for implementation in general-purpose imageprocessing.and image analysis pipeline architectures. Also, some new pipeline configurations are proposed that achieve a remarkable degree of parallelism in the computation of projection data and, in fact, of many other geometrical descriptors of digitalimages. In particular, random access memories and other dedicated hardware devices are not needed in these algorithms. The effectiveness of the approach and feasibility of the proposed architectures are demonstrated by running the algorithms in commercially available shortpipelines for imageprocessing.and analysis. Examples are shown of the use of projection data for machine vision applications.
This paper presents a new imageprocessing.system intends to enhance video images transmitted from remotely piloted vehicles (R. P. Vs), or any other airphotoes in real time. The objective was to select known algorith...
详细信息
ISBN:
(纸本)0892526149
This paper presents a new imageprocessing.system intends to enhance video images transmitted from remotely piloted vehicles (R. P. Vs), or any other airphotoes in real time. The objective was to select known algorithms of imageprocessing.techniques and to develop special purpose architecture, based on pipe line processor in hardware and top down structure design in software, in order to apply these algorithms in real time. The system operator is not required to have knowledge of imageprocessing. Interactive operations were chosen since successful image enhancement is based on the ability of the observer to detect information of interest in the final process image. The system architecture, the imageprocessing.moduls and the data base management are described briefly.
This conference proceedings contains 14 papers, two of which are in abstract form only. The main subjects are local neighborhood operators, display of digitalimage data, remote sensing, image thinning algorithms, rea...
详细信息
ISBN:
(纸本)0869804731
This conference proceedings contains 14 papers, two of which are in abstract form only. The main subjects are local neighborhood operators, display of digitalimage data, remote sensing, image thinning algorithms, realtime data compression, syntactic edge detection algorithms, parallel processing.image computers, computer vision systems, shape recognition, and characterization of visual texture.
The proceedings contain 22 papers. The topics discussed include: a dualistic model to describe computer architectures;modular real time architectures for image understanding applications;a high speed recirculating nei...
The proceedings contain 22 papers. The topics discussed include: a dualistic model to describe computer architectures;modular real time architectures for image understanding applications;a high speed recirculating neighborhood processing.architecture;system design for local neighborhood processing.the design of video processor in TV tracking system;a new architecture for real-time imageprocessing.advanced digital video processing.unit;obtaining feasible solutions in image restoration;a study of the discrete Hartley transform for image compression applications;block transform image compression by power spectrum fitting;and superfine laser position control using statistically enhanced resolution in real time.
Low level processing.or preprocessing.is often the starting point in many computer vision systems. This stage is typically characterized as picture processing.and uses gradients, edge templates, filters, thresholding,...
详细信息
ISBN:
(纸本)0892526149
Low level processing.or preprocessing.is often the starting point in many computer vision systems. This stage is typically characterized as picture processing.and uses gradients, edge templates, filters, thresholding, and other enhancement techniques. The next stage in computer vision commonly deals with the processed image, usually binary in nature, and extracts intrinsic features such as size, shape, boundary, orientation, moments, etc. A new class of multiple-valued integrated optical processors is proposed for the parallel implementation of selected preprocessing.algorithms. The basic component of the multiple-valued processor is an integrated optical threshold gate. The parallel processing.algorithms required for both smoothing and boundary determination with an integrated optical thresholding array are described. One possible architecture for an optical parallel digitalimage processor is presented.
The project described is concerned with interfacing between the computer knowledge based system and the real world in the areas of speech recognition and image recognition and understanding. At the upper level the sys...
详细信息
ISBN:
(纸本)0444877401
The project described is concerned with interfacing between the computer knowledge based system and the real world in the areas of speech recognition and image recognition and understanding. At the upper level the system will consist of knowledge processing.while the lower levels will consist of preprocessing.and feature extraction. The paper discusses hardware and algorithmic aspects of the system. The systems need to operate in near real time and so providing processing.power is a key feature of the work.
Several factors which contribute to the degradation of SLAM imagery are identified and discussed. These include (1) variable optical intensity due to both laser intensity fluctuations and spatially varying sample or c...
详细信息
Several factors which contribute to the degradation of SLAM imagery are identified and discussed. These include (1) variable optical intensity due to both laser intensity fluctuations and spatially varying sample or coverslip reflectivity and (2) the mixing of the elastic response of the sample surface and cover slip with the impinging elastic wavefronts which contain the desired image information. A computer-controlled SLAM and several spatial frequency filtering algorithms are described that significantly mitigate these problems. Several images obtained from test samples are presented that illustrate these points. The digitally processed SLAM images are comparable in quality to those that a scanning acoustic microscopy operating at the same acoustic frequency (100 MHz in this study) could produce.
暂无评论