Summary form only given. A fast edge detector architecture and IC, based on a new edge follower algorithm, have been designed. The chip offers real-time processing with a limited amount of hardware due to the optimization of the critical path in the architecture. In this way, a complete frame (512*512) can be processed in about 400000 clock cycles, and a clock rate of up to 10 MHz has been achieved in a 3- mu m double-metal CMOS technology. This chip offers online information such as edge location and orientation, which can be used for feature extraction and pattern recognition in the robot vision system. A novel architectural model, the multiplexed cooperating datapath architecture, has been adopted to obtain an efficient design with a minimal set of functional building blocks. The methodology is especially suited for recursive types of algorithms such as the edge follower. High throughput is achieved by optimizing the memory storage and by eliminating the communication bottlenecks with dedicated buses.<>
{"title":"A fast edge detection chip for robot vision systems","authors":"C.Y. Lee, F. Catthoor, H. de Man","doi":"10.1109/MDSP.1989.97027","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97027","url":null,"abstract":"Summary form only given. A fast edge detector architecture and IC, based on a new edge follower algorithm, have been designed. The chip offers real-time processing with a limited amount of hardware due to the optimization of the critical path in the architecture. In this way, a complete frame (512*512) can be processed in about 400000 clock cycles, and a clock rate of up to 10 MHz has been achieved in a 3- mu m double-metal CMOS technology. This chip offers online information such as edge location and orientation, which can be used for feature extraction and pattern recognition in the robot vision system. A novel architectural model, the multiplexed cooperating datapath architecture, has been adopted to obtain an efficient design with a minimal set of functional building blocks. The methodology is especially suited for recursive types of algorithms such as the edge follower. High throughput is achieved by optimizing the memory storage and by eliminating the communication bottlenecks with dedicated buses.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116314270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. The Frazier-Jawerth transform (FJT), originally the phi-transform, is similar to the wavelet transform and is distinguished by the fact that the analyzing functions form an overcomplete basis for he signal space and may be nonorthogonal. This added flexibility makes possible the definition of optimal analyzing functions, which are the focus of this study. For continuous-time and infinite discrete-time signals, the optimally localized functions are the prolate spheroidal wave functions and their discrete versions. For finite discrete-time signals and images, generalizations of these functions that are applicable for use in the FJT have been identified by the authors.<>
{"title":"Compact functions and the Frazier-Jawerth transform","authors":"D. Fuhrmann, A. Kumar, J. R. Cox","doi":"10.1109/MDSP.1989.97063","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97063","url":null,"abstract":"Summary form only given. The Frazier-Jawerth transform (FJT), originally the phi-transform, is similar to the wavelet transform and is distinguished by the fact that the analyzing functions form an overcomplete basis for he signal space and may be nonorthogonal. This added flexibility makes possible the definition of optimal analyzing functions, which are the focus of this study. For continuous-time and infinite discrete-time signals, the optimally localized functions are the prolate spheroidal wave functions and their discrete versions. For finite discrete-time signals and images, generalizations of these functions that are applicable for use in the FJT have been identified by the authors.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"172 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121034439","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. The problems of determining the location of the epipoles of two displaced image planes when they are displaced due either to the motion of an object along the time axis or to the displacement of stereo cameras along the space axes have been investigated. It has been assumed that a rigid object moves with respect to still stereo camera systems.<>
{"title":"Motion parameter estimation-from spatial-temporal matching","authors":"C. Chang, S. Chatterjee","doi":"10.1109/MDSP.1989.96988","DOIUrl":"https://doi.org/10.1109/MDSP.1989.96988","url":null,"abstract":"Summary form only given. The problems of determining the location of the epipoles of two displaced image planes when they are displaced due either to the motion of an object along the time axis or to the displacement of stereo cameras along the space axes have been investigated. It has been assumed that a rigid object moves with respect to still stereo camera systems.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"101 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121590014","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. A novel approach to the design of perfect-reconstruction 2D mirror image filter banks, using auxiliary channel(s), is proposed. It is assumed that the filters are separable and identical in both directions. It has been shown that the structure can provide perfect reconstruction.<>
{"title":"A new approach to the design of perfect reconstruction two-dimensional mirror image filter bank using an auxiliary channel","authors":"O. Johnsen, H. Babic, S. Mitra, O. Shentov","doi":"10.1109/MDSP.1989.97141","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97141","url":null,"abstract":"Summary form only given. A novel approach to the design of perfect-reconstruction 2D mirror image filter banks, using auxiliary channel(s), is proposed. It is assumed that the filters are separable and identical in both directions. It has been shown that the structure can provide perfect reconstruction.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121904832","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given, as follows. In certain circumstances, it is not possible to improve radar detection performance using conventional radar techniques, i.e. increased power, shorter pulse length, coherence, etc. If the radar polarization characteristics of the target are sufficiently different from those of the surrounding clutter environment, it is possible to improve detection through the use of polarization-domain processing. The polarization state (PS) can be viewed as adding new dimensions to the conventional 1-D echo amplitude normally used for detection. The application of a multidimensional, multichannel distance metric in the amplitude and polarization domains to detect cooperative retroreflectors with distinctive radar polarization characteristics in man-made and natural clutter environments is described. The results are based on real data collected using a partially coherent X-band weather radar system.<>
{"title":"Radar detection of co-operative targets using dual polarized radar-a multidimensional, multichannel detection problem","authors":"A. Macikunas, S. Haykin","doi":"10.1109/MDSP.1989.97010","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97010","url":null,"abstract":"Summary form only given, as follows. In certain circumstances, it is not possible to improve radar detection performance using conventional radar techniques, i.e. increased power, shorter pulse length, coherence, etc. If the radar polarization characteristics of the target are sufficiently different from those of the surrounding clutter environment, it is possible to improve detection through the use of polarization-domain processing. The polarization state (PS) can be viewed as adding new dimensions to the conventional 1-D echo amplitude normally used for detection. The application of a multidimensional, multichannel distance metric in the amplitude and polarization domains to detect cooperative retroreflectors with distinctive radar polarization characteristics in man-made and natural clutter environments is described. The results are based on real data collected using a partially coherent X-band weather radar system.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126181366","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given, as follows. A machine vision system has been designed and constructed for automatic visual measurement of wear patterns on metal cutting tools used for machine shop lathes. Applications of the system include (1) adaptive control of the tool cutting path based on the tool wear to ensure high-precision machining and (2) prediction of tool breakage determined from wear patterns and cutting force measurements. Phase-stepping interferometry is used to project sinusoidal patterns onto the reflective surfaces of the cutting tool. Using multiple phase shifts of the patterns, three-dimensional information about the edge of the tool can be extracted from video images. The spatial resolution of the system is on the order of 2 to 5 mu m. A description of the system together with experimental results will be given.<>
{"title":"Visual measurement of metal cutting tool wear","authors":"D. Capson, C. Wust","doi":"10.1109/MDSP.1989.97029","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97029","url":null,"abstract":"Summary form only given, as follows. A machine vision system has been designed and constructed for automatic visual measurement of wear patterns on metal cutting tools used for machine shop lathes. Applications of the system include (1) adaptive control of the tool cutting path based on the tool wear to ensure high-precision machining and (2) prediction of tool breakage determined from wear patterns and cutting force measurements. Phase-stepping interferometry is used to project sinusoidal patterns onto the reflective surfaces of the cutting tool. Using multiple phase shifts of the patterns, three-dimensional information about the edge of the tool can be extracted from video images. The spatial resolution of the system is on the order of 2 to 5 mu m. A description of the system together with experimental results will be given.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125970019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The specific problem that was addressed is one in which there is limited data in both the temporal and spatial dimensions, so that one cannot assume the use of ordinary Fourier transforms on the time domain outputs of each sensor. Rather, zero-mean Gaussian statistics were assumed, and the likelihood of the observed data was directly maximized with respect to the parameters which enter into the covariance matrix of the multivariate distribution. Two models were pursued. The first is a parametric model in which it is assumed that there are a fixed number of independent, wide-sense-stationary, plane-wave signals. The second model is one in which there is energy impinging upon the array from a spatial continuum. EM (expectation-maximization) algorithms appropriate for these two problems were derived.<>
{"title":"Maximum-likelihood wideband direction-of-arrival estimation","authors":"D. Fuhrmann, M. Miller","doi":"10.1109/MDSP.1989.97076","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97076","url":null,"abstract":"The specific problem that was addressed is one in which there is limited data in both the temporal and spatial dimensions, so that one cannot assume the use of ordinary Fourier transforms on the time domain outputs of each sensor. Rather, zero-mean Gaussian statistics were assumed, and the likelihood of the observed data was directly maximized with respect to the parameters which enter into the covariance matrix of the multivariate distribution. Two models were pursued. The first is a parametric model in which it is assumed that there are a fixed number of independent, wide-sense-stationary, plane-wave signals. The second model is one in which there is energy impinging upon the array from a spatial continuum. EM (expectation-maximization) algorithms appropriate for these two problems were derived.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122524838","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. The superresolution estimation of bearings of wideband sources using fewer sensors than sources was addressed. The fundamental limits on the maximum number of wideband sources uniquely resolvable by a passive sensor array, called the resolution capacity of the array, were studied. The source signals are assumed to be uncorrelated zero-mean, second-order stationary, and ergodic wideband random processes occupying a common bandwidth. The array is assumed to be uniform and linear, and L>>1 Nyquist-rate samples of its output are assumed to be collected into a snapshot supervector. N>>1 such snapshots are assumed to be available.<>
{"title":"On the resolution capacity of wideband sensor arrays","authors":"Y. Bresler, A.J. Ficker","doi":"10.1109/MDSP.1989.97073","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97073","url":null,"abstract":"Summary form only given. The superresolution estimation of bearings of wideband sources using fewer sensors than sources was addressed. The fundamental limits on the maximum number of wideband sources uniquely resolvable by a passive sensor array, called the resolution capacity of the array, were studied. The source signals are assumed to be uncorrelated zero-mean, second-order stationary, and ergodic wideband random processes occupying a common bandwidth. The array is assumed to be uniform and linear, and L>>1 Nyquist-rate samples of its output are assumed to be collected into a snapshot supervector. N>>1 such snapshots are assumed to be available.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130560943","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. It has been shown that the problem of detecting edges in a digital image is equivalent to the problem of estimating the wave number vectors of complex exponentials in the spatial frequency domain. This observation has been used to show that most of the known non-model-based edge detection algorithms can be interpreted as variations of the periodogram method of spectral estimation. The above observation has also been used to derive three edge detection algorithms. The first algorithm is based on the fact that complex exponentials are the homogeneous solution of a difference equation with proper initial conditions. It derives estimates of the edge locations by performing a singular-value decomposition of a Hankel matrix formed from the fast Fourier transform of the underlying image. The second and third approaches use the maximum-likelihood spectral estimation method and various maximum-entropy spectral estimation technique on the fast Fourier transform of the underlying image to estimate the edge locations. The main advantage of the three approaches is that they do not involve the use of a smoothing filter or gradient operations.<>
{"title":"Edge detection using spectral estimation techniques","authors":"A. Tewfik, F. Assaad, Mohamed Deriche","doi":"10.1109/MDSP.1989.97006","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97006","url":null,"abstract":"Summary form only given. It has been shown that the problem of detecting edges in a digital image is equivalent to the problem of estimating the wave number vectors of complex exponentials in the spatial frequency domain. This observation has been used to show that most of the known non-model-based edge detection algorithms can be interpreted as variations of the periodogram method of spectral estimation. The above observation has also been used to derive three edge detection algorithms. The first algorithm is based on the fact that complex exponentials are the homogeneous solution of a difference equation with proper initial conditions. It derives estimates of the edge locations by performing a singular-value decomposition of a Hankel matrix formed from the fast Fourier transform of the underlying image. The second and third approaches use the maximum-likelihood spectral estimation method and various maximum-entropy spectral estimation technique on the fast Fourier transform of the underlying image to estimate the edge locations. The main advantage of the three approaches is that they do not involve the use of a smoothing filter or gradient operations.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131984854","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Summary form only given. Simultaneous iterative identification and restoration have been treated. The image and the noise have been modeled as multivariate Gaussian processes. Maximum-likelihood estimation has been used to estimate the parameters that characterize the Gaussian processes, where the estimation of the conditional mean of the image represents the restored image. Likelihood functions of observed images are highly nonlinear with respect to these parameters. Therefore, it is in general very difficult to maximize them directly. The expectation-maximization (EM) algorithm has been used to find these parameters.<>
{"title":"Maximum likelihood image identification and restoration based on the EM algorithm","authors":"A. Katsaggelos","doi":"10.1109/MDSP.1989.97107","DOIUrl":"https://doi.org/10.1109/MDSP.1989.97107","url":null,"abstract":"Summary form only given. Simultaneous iterative identification and restoration have been treated. The image and the noise have been modeled as multivariate Gaussian processes. Maximum-likelihood estimation has been used to estimate the parameters that characterize the Gaussian processes, where the estimation of the conditional mean of the image represents the restored image. Likelihood functions of observed images are highly nonlinear with respect to these parameters. Therefore, it is in general very difficult to maximize them directly. The expectation-maximization (EM) algorithm has been used to find these parameters.<<ETX>>","PeriodicalId":340681,"journal":{"name":"Sixth Multidimensional Signal Processing Workshop,","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1989-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134182235","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：