Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770776
A. Košir, J. Tasic
Pattern spectrum is a shape size descriptor of the object content of a digital image. It is a technique for extracting shape and resolution information of a digital image. A discussion of the basic ideas of pattern spectrum is presented in the paper. Some useful features of the pattern spectrum concerning invariants for object recognition are derived. Features important for the calculation of the pattern spectrum are also included. The technique is applied to some typical shapes.
{"title":"Pattern spectrum of binary image","authors":"A. Košir, J. Tasic","doi":"10.1109/ICOSP.1998.770776","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770776","url":null,"abstract":"Pattern spectrum is a shape size descriptor of the object content of a digital image. It is a technique for extracting shape and resolution information of a digital image. A discussion of the basic ideas of pattern spectrum is presented in the paper. Some useful features of the pattern spectrum concerning invariants for object recognition are derived. Features important for the calculation of the pattern spectrum are also included. The technique is applied to some typical shapes.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124709023","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}
Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770820
Guojun Lu, J. Phillips
In common colour-based image retrieval, colour histograms for images in the database and queries are calculated. The distance between a query and each of the database images is calculated as the sum of the absolute values of bin-to-bin differences between their histograms. The method ignores colour similarity between bins, leading to cases where perceptually similar images have very large histogram distances. In this paper, we propose to use perceptually weighted histograms (PWH) to overcome the problem. In PWH, a pixel contributes weights to a number of perceptually similar bins instead of a single bin. The contributing weights are inversely proportional to the distance between the pixel and bins.
{"title":"Using perceptually weighted histograms for colour-based image retrieval","authors":"Guojun Lu, J. Phillips","doi":"10.1109/ICOSP.1998.770820","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770820","url":null,"abstract":"In common colour-based image retrieval, colour histograms for images in the database and queries are calculated. The distance between a query and each of the database images is calculated as the sum of the absolute values of bin-to-bin differences between their histograms. The method ignores colour similarity between bins, leading to cases where perceptually similar images have very large histogram distances. In this paper, we propose to use perceptually weighted histograms (PWH) to overcome the problem. In PWH, a pixel contributes weights to a number of perceptually similar bins instead of a single bin. The contributing weights are inversely proportional to the distance between the pixel and bins.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121899348","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}
Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770930
F. Gao, D. Shattuck
In recent years,the possibility of resistivity logging in cased wells has attracted much interest. This study draws on the basic principles of through-casing resistivity (TCR) measurements. Based on the principle of TCR theory, we designed a TCR tool and performed TCR experiments. Computer simulation code was written to provide expected values. In the experiments, we used two formation beds with a high contrast in resistivity, an airbed and a saline waterbed. We measured two sets of logs, which gave distinguishable readings between air and saline water. The measured results were compared to values predicted by the computer model. We found that the distribution of the leakage current was not uniform in our experiment, which was unexpected. However, these results can be used in the future to improve TCR measurements and interpretation.
{"title":"A scale model of the through-casing resistivity measurement","authors":"F. Gao, D. Shattuck","doi":"10.1109/ICOSP.1998.770930","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770930","url":null,"abstract":"In recent years,the possibility of resistivity logging in cased wells has attracted much interest. This study draws on the basic principles of through-casing resistivity (TCR) measurements. Based on the principle of TCR theory, we designed a TCR tool and performed TCR experiments. Computer simulation code was written to provide expected values. In the experiments, we used two formation beds with a high contrast in resistivity, an airbed and a saline waterbed. We measured two sets of logs, which gave distinguishable readings between air and saline water. The measured results were compared to values predicted by the computer model. We found that the distribution of the leakage current was not uniform in our experiment, which was unexpected. However, these results can be used in the future to improve TCR measurements and interpretation.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124021855","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}
Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770236
Peng Xinguang, Jiang Zhongning
A linear processor array interconnected by optical fiber buses is presented in the paper. It is especially suitable for application such as intensive signal processing and intensive communication. The signals can be transmitted concurrently by any different processor, in a pipeline fashion on the optical fiber buses without collision of message data. The linear processor array incorporates high-speed fiber communication with high-performance computing of processors. A new technique of optical pulse coincident addressing and a parallel algorithm for the solution of nonlinear equations are simultaneously proposed in the paper.
{"title":"A linear processor array used for intensive communication","authors":"Peng Xinguang, Jiang Zhongning","doi":"10.1109/ICOSP.1998.770236","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770236","url":null,"abstract":"A linear processor array interconnected by optical fiber buses is presented in the paper. It is especially suitable for application such as intensive signal processing and intensive communication. The signals can be transmitted concurrently by any different processor, in a pipeline fashion on the optical fiber buses without collision of message data. The linear processor array incorporates high-speed fiber communication with high-performance computing of processors. A new technique of optical pulse coincident addressing and a parallel algorithm for the solution of nonlinear equations are simultaneously proposed in the paper.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125785394","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}
Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770309
T. Reynolds, E. Pizzolato
The multinet phone classifier architecture is a framework for combining specialised phone detection networks into a posterior probability estimator for all phones. In this paper we give results obtained for the architecture on TIMIT phone classification tasks. We compare it with a standard mixture of Gaussian HMM classifiers.
{"title":"Phoneme classification with multinets","authors":"T. Reynolds, E. Pizzolato","doi":"10.1109/ICOSP.1998.770309","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770309","url":null,"abstract":"The multinet phone classifier architecture is a framework for combining specialised phone detection networks into a posterior probability estimator for all phones. In this paper we give results obtained for the architecture on TIMIT phone classification tasks. We compare it with a standard mixture of Gaussian HMM classifiers.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130105865","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}
Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770202
Jun Wang, Hongwei Liu, Shouhong Zhang
Two typical types of signal model known as narrowband and wideband signal models are frequently used in radar, sonar and communication signal processing. A matched filter in two-dimensional linear time-frequency plane is studied and the optimal signal detection methods based on match projection onto time-frequency and time-scale subspace are presented for the effective detection of these two-type signals. Simulation results show that the methods have good performance and robustness.
{"title":"Signal detection based on match projection onto time-frequency and time-scale subspace","authors":"Jun Wang, Hongwei Liu, Shouhong Zhang","doi":"10.1109/ICOSP.1998.770202","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770202","url":null,"abstract":"Two typical types of signal model known as narrowband and wideband signal models are frequently used in radar, sonar and communication signal processing. A matched filter in two-dimensional linear time-frequency plane is studied and the optimal signal detection methods based on match projection onto time-frequency and time-scale subspace are presented for the effective detection of these two-type signals. Simulation results show that the methods have good performance and robustness.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"1940 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129288242","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}
Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770247
W. Zheng
This paper presents a new type of improved least-squares (ILS) algorithm for adaptive parameter estimation of autoregressive (AR) signals from noisy observations. Unlike the previous ILS-based methods, the developed algorithm can give consistent parameter estimates in a very direct manner that does not involve dealing with an augmented noisy AR model. The new algorithm is demonstrated to outperform the previous ILS-based methods in terms of its improved numerical efficiency.
{"title":"Adaptive parameter estimation of autoregressive signals from noisy observations","authors":"W. Zheng","doi":"10.1109/ICOSP.1998.770247","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770247","url":null,"abstract":"This paper presents a new type of improved least-squares (ILS) algorithm for adaptive parameter estimation of autoregressive (AR) signals from noisy observations. Unlike the previous ILS-based methods, the developed algorithm can give consistent parameter estimates in a very direct manner that does not involve dealing with an augmented noisy AR model. The new algorithm is demonstrated to outperform the previous ILS-based methods in terms of its improved numerical efficiency.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129564066","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}
Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770822
Zhang Ying, Tao Ran, Z. Siyong, Wang Yue
Multimedia technologies have brought a revolution in information science. As the main technology for the broadband integrated services network, asynchronous transfer mode (ATM) provides efficient switching capability for multimedia communication services as well as service quality guarantees. Quality of service (QoS) estimation is one of the key issues in traffic engineering. A new approach to cell loss ratio (CLR) prediction and estimation is presented in this paper, which is based on simple source traffic models constructed with standard parameters defined by the ITU and ATM Forum. This method can estimate the burst level CLR more efficiently than typical analytical approaches while maintaining accurate enough results to be used in real-time applications. Description of the algorithm, some numerical results and comparison with simulations are also included to show its simplicity and efficiency.
{"title":"A new algorithm for real-time QoS estimation in multimedia communication systems","authors":"Zhang Ying, Tao Ran, Z. Siyong, Wang Yue","doi":"10.1109/ICOSP.1998.770822","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770822","url":null,"abstract":"Multimedia technologies have brought a revolution in information science. As the main technology for the broadband integrated services network, asynchronous transfer mode (ATM) provides efficient switching capability for multimedia communication services as well as service quality guarantees. Quality of service (QoS) estimation is one of the key issues in traffic engineering. A new approach to cell loss ratio (CLR) prediction and estimation is presented in this paper, which is based on simple source traffic models constructed with standard parameters defined by the ITU and ATM Forum. This method can estimate the burst level CLR more efficiently than typical analytical approaches while maintaining accurate enough results to be used in real-time applications. Description of the algorithm, some numerical results and comparison with simulations are also included to show its simplicity and efficiency.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128434868","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}
Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770156
J. Gustavsson, Sven Nordebo, P. Börjesson
This paper treats channel estimation and signal detection in Laplacian noise. The received signal is assumed to be a transmitted signal which has been corrupted by an unknown channel, modeled as a FIR filter, the output being further disturbed by additive independent Laplacian noise. The transmitted signal is assumed to depend on an unknown parameter belonging to a known finite set. The simultaneous maximum likelihood (ML) estimator of the unknown parameter, as well as of the FIR filter coefficients, is derived. The ML estimate of the channel can be obtained by using a linear programming approach and the decision about the parameter is based on the output from a set of generalized matched filters. Simulation results are included in order to illustrate the performance of the proposed receivers.
{"title":"Simultaneous channel and symbol maximum likelihood estimation in Laplacian noise","authors":"J. Gustavsson, Sven Nordebo, P. Börjesson","doi":"10.1109/ICOSP.1998.770156","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770156","url":null,"abstract":"This paper treats channel estimation and signal detection in Laplacian noise. The received signal is assumed to be a transmitted signal which has been corrupted by an unknown channel, modeled as a FIR filter, the output being further disturbed by additive independent Laplacian noise. The transmitted signal is assumed to depend on an unknown parameter belonging to a known finite set. The simultaneous maximum likelihood (ML) estimator of the unknown parameter, as well as of the FIR filter coefficients, is derived. The ML estimate of the channel can be obtained by using a linear programming approach and the decision about the parameter is based on the output from a set of generalized matched filters. Simulation results are included in order to illustrate the performance of the proposed receivers.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128556230","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}
Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770865
De-shuang Huang
This paper discusses the relationship between linear feedforward neural network classifiers (FNNC) and the reduced-rank approximation. From the viewpoint of linear algebra, it is shown that if the rank of the trained connection weight matrix of a two layered linear FNNC is greater than or equal to the rank of the between-class dispersion matrix of the input training samples, the two layered linear FNNC will be merged into a one layered linear FNNC. In addition, the condition of the null error cost function for a reduced rank approximation is also derived.
{"title":"Linear feedforward neural network classifiers and reduced-rank approximation","authors":"De-shuang Huang","doi":"10.1109/ICOSP.1998.770865","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770865","url":null,"abstract":"This paper discusses the relationship between linear feedforward neural network classifiers (FNNC) and the reduced-rank approximation. From the viewpoint of linear algebra, it is shown that if the rank of the trained connection weight matrix of a two layered linear FNNC is greater than or equal to the rank of the between-class dispersion matrix of the input training samples, the two layered linear FNNC will be merged into a one layered linear FNNC. In addition, the condition of the null error cost function for a reduced rank approximation is also derived.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"97 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120885156","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：