Pub Date : 1998-10-12DOI: 10.1109/ICOSP.1998.770245
Hyoung-Nam Kim, Woo‐Jin Song
An adaptive IIR equalization method for nonminimum-phase channels is presented. In order to ensure the stability of the IIR equalizer, we propose a novel way to obtain a minimum-phase equalized channel from a nonminimum-phase channel. A FIR direct equalizer is followed by an all-pole IIR indirect equalizer to form an optimal equalizer structure. The minimum-phase equalized channel is directly inverted by the indirect equalizer and the remaining part of the channel is handled by the direct equalizer. The Routh-Hurwitz criterion is introduced for verifying that the directly-equalized channel is minimum phase. With the same number of taps, the proposed IIR equalizer performs much better than conventional FIR equalizers in terms of residual MSE.
{"title":"An adaptive IIR equalizer for nonminimum-phase channels","authors":"Hyoung-Nam Kim, Woo‐Jin Song","doi":"10.1109/ICOSP.1998.770245","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770245","url":null,"abstract":"An adaptive IIR equalization method for nonminimum-phase channels is presented. In order to ensure the stability of the IIR equalizer, we propose a novel way to obtain a minimum-phase equalized channel from a nonminimum-phase channel. A FIR direct equalizer is followed by an all-pole IIR indirect equalizer to form an optimal equalizer structure. The minimum-phase equalized channel is directly inverted by the indirect equalizer and the remaining part of the channel is handled by the direct equalizer. The Routh-Hurwitz criterion is introduced for verifying that the directly-equalized channel is minimum phase. With the same number of taps, the proposed IIR equalizer performs much better than conventional FIR equalizers in terms of residual MSE.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"26 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":"116859899","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.1998.770339
Zhang Xudong, Wang Desheng, Peng Ying-ning
A novel video coding scheme based on the blocking regional wavelet transform and overlapped motion compensation is researched in this paper. Several new or improved algorithms are proposed. The high-energy regions of interframe predicted difference are first detected using a hierarchical sliding-window region forming algorithm. The regional border is processed to reduce ringing-effects. Wavelet decomposition and quantization are carried out only for these regions. A quantization model based on HVS spatiality-frequency and time-frequency features is used in the coding algorithm. An improved single-pass scanned zero-tree is employed to represent quantized wavelet coefficients. This scheme reduces the bit rate by about 36%-51% or gets a distinct improvement in decoding video quality by cutting down ringing-effects at the same bit rate compared with traditional wavelet-based coding algorithms. Simulation results show acceptable image quality for CIF resolution at low bit-rate and very low bit-rate.
{"title":"A low bit-rate video coding algorithm based on blocking regional wavelet transform with border processing","authors":"Zhang Xudong, Wang Desheng, Peng Ying-ning","doi":"10.1109/ICOSP.1998.770339","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770339","url":null,"abstract":"A novel video coding scheme based on the blocking regional wavelet transform and overlapped motion compensation is researched in this paper. Several new or improved algorithms are proposed. The high-energy regions of interframe predicted difference are first detected using a hierarchical sliding-window region forming algorithm. The regional border is processed to reduce ringing-effects. Wavelet decomposition and quantization are carried out only for these regions. A quantization model based on HVS spatiality-frequency and time-frequency features is used in the coding algorithm. An improved single-pass scanned zero-tree is employed to represent quantized wavelet coefficients. This scheme reduces the bit rate by about 36%-51% or gets a distinct improvement in decoding video quality by cutting down ringing-effects at the same bit rate compared with traditional wavelet-based coding algorithms. Simulation results show acceptable image quality for CIF resolution at low bit-rate and very low bit-rate.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"82 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":"123928374","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.1998.770778
Hao Shi
The ever-increasing amount of information available to meteorologists has led to a better understanding of weather forecasting. The ability to automatically detect and estimate cloud segment speed and direction in a weather satellite would be very useful in forecasts of future weather events. The objective of this research is to extract a particular cloud segment from a sequence of infrared satellite images, and detect cloud movement and calculate its speed and direction. The difference centroid algorithm is employed to detect the movement of the cloud segment. Cloud speed and direction are then calculated by finding centroid displacement between cloud segments.
{"title":"Cloud movement detection for satellite images","authors":"Hao Shi","doi":"10.1109/ICOSP.1998.770778","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770778","url":null,"abstract":"The ever-increasing amount of information available to meteorologists has led to a better understanding of weather forecasting. The ability to automatically detect and estimate cloud segment speed and direction in a weather satellite would be very useful in forecasts of future weather events. The objective of this research is to extract a particular cloud segment from a sequence of infrared satellite images, and detect cloud movement and calculate its speed and direction. The difference centroid algorithm is employed to detect the movement of the cloud segment. Cloud speed and direction are then calculated by finding centroid displacement between cloud segments.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"10 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":"129719638","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.1998.770828
Yanning Zhang, L. Jiao, Hu Songhua
The People's Republic of China's fishery and offshore petroleum development industries have been in urgent need of a classifier of noise signals. In this paper, a local adaptive wavelet neural network is proposed, and an efficient engineering classifier based on the local adaptive wavelet neural network is designed and applied to classifying actual ship noises. The classification experiment results are encouraging, which shows that the classifier above is an efficient engineering classifier for actual ship noises.
{"title":"An efficient method of target classification","authors":"Yanning Zhang, L. Jiao, Hu Songhua","doi":"10.1109/ICOSP.1998.770828","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770828","url":null,"abstract":"The People's Republic of China's fishery and offshore petroleum development industries have been in urgent need of a classifier of noise signals. In this paper, a local adaptive wavelet neural network is proposed, and an efficient engineering classifier based on the local adaptive wavelet neural network is designed and applied to classifying actual ship noises. The classification experiment results are encouraging, which shows that the classifier above is an efficient engineering classifier for actual ship noises.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"67 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":"128318818","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.1998.770795
H. Kuroda, Tianxing Liu
We present a picture compression algorithm based on linear transformation, vector space and matrix algebra. We consider the picture as the matrix in this algorithm. The picture compression is realized by means of simplifying the picture matrix. To achieve this aim, the vector groups in the picture submatrix are restructured according to the relation of linear dependence or linear independence among them. After the keyword parameters of the algorithm are adjusted, it is fit for compressing static and dynamic pictures.
{"title":"An image compressing algorithm based on vector space and algebra of matrix","authors":"H. Kuroda, Tianxing Liu","doi":"10.1109/ICOSP.1998.770795","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770795","url":null,"abstract":"We present a picture compression algorithm based on linear transformation, vector space and matrix algebra. We consider the picture as the matrix in this algorithm. The picture compression is realized by means of simplifying the picture matrix. To achieve this aim, the vector groups in the picture submatrix are restructured according to the relation of linear dependence or linear independence among them. After the keyword parameters of the algorithm are adjusted, it is fit for compressing static and dynamic pictures.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"2016 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":"128262571","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.1998.770863
Z. Dongming, Cao Jinde
This paper studies the problem of global exponential stability of the equilibrium of a class of neural networks with delays, by using the method of variation of constants and combining it with the method of inequality analysis. Sufficient conditions for global exponential stability of neural networks with delays are obtained, for which we do not require symmetry of the connection matrix and nonlinear properties for neural units to be continuously differentiable or strictly monotonically increasing. These conditions can be used to design globally stable neural networks.
{"title":"Global exponential stability in neural network with delays","authors":"Z. Dongming, Cao Jinde","doi":"10.1109/ICOSP.1998.770863","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770863","url":null,"abstract":"This paper studies the problem of global exponential stability of the equilibrium of a class of neural networks with delays, by using the method of variation of constants and combining it with the method of inequality analysis. Sufficient conditions for global exponential stability of neural networks with delays are obtained, for which we do not require symmetry of the connection matrix and nonlinear properties for neural units to be continuously differentiable or strictly monotonically increasing. These conditions can be used to design globally stable neural networks.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"2 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":"129668505","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.1998.770170
R. C. Papademetriou
The use of the minimum cross-entropy (MCE) principle in spectrum estimation, resulting in an information-theoretic method that explicitly includes prior spectral information, has followed two philosophically different pathways leading to two distinct estimators, cross-entropy (CE) and spectral cross-entropy (SCE) respectively. These estimators are compared experimentally, regarding resolvability and fidelity. The data model assumed consists of two equal-amplitude sinusoidal signals, immersed in 1/f noise. The CE estimator appears to have evident superiority in this study.
{"title":"Experimental comparison of two information-theoretic spectral estimators","authors":"R. C. Papademetriou","doi":"10.1109/ICOSP.1998.770170","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770170","url":null,"abstract":"The use of the minimum cross-entropy (MCE) principle in spectrum estimation, resulting in an information-theoretic method that explicitly includes prior spectral information, has followed two philosophically different pathways leading to two distinct estimators, cross-entropy (CE) and spectral cross-entropy (SCE) respectively. These estimators are compared experimentally, regarding resolvability and fidelity. The data model assumed consists of two equal-amplitude sinusoidal signals, immersed in 1/f noise. The CE estimator appears to have evident superiority in this study.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"116 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":"128906923","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.1998.770167
Wang Huakui, Z. Liyi, L. Ruilian
This paper presents a nonlinear adaptive recursive digital filter in which an adaptively adjustable nonlinear function is placed at the output of a linear recursive filter. Adaptive algorithms are based on a gradient search technique for updating coefficients of the filter and parameters of the nonlinear function. The filter is stable in the sense that the filter output is always bounded with bounded input signals, and it does not need stability monitoring in the adaptive process. Computer simulations confirm the validity of the derived results.
{"title":"Nonlinear adaptive recursive filtering using gradient-based algorithms","authors":"Wang Huakui, Z. Liyi, L. Ruilian","doi":"10.1109/ICOSP.1998.770167","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770167","url":null,"abstract":"This paper presents a nonlinear adaptive recursive digital filter in which an adaptively adjustable nonlinear function is placed at the output of a linear recursive filter. Adaptive algorithms are based on a gradient search technique for updating coefficients of the filter and parameters of the nonlinear function. The filter is stable in the sense that the filter output is always bounded with bounded input signals, and it does not need stability monitoring in the adaptive process. Computer simulations confirm the validity of the derived results.","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":"130649807","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.1998.770200
Peng Zhiwei, Wang Bo, Liao Guisheng
Compactly supported orthogonal wavelets are obtained from two-band paraunitary FIR filter bank solutions, with the additional "flatness" constraint that the low-pass filter should have K zeros at half of the sampling frequency. This constraint is set to obtain a "regular" wavelet, but it is somewhat in contradiction with the requirement of excellent frequency selectivity. Maximally flat filters yielding Daubechies wavelets have the worst frequency selectivity. According to those drawbacks, two design methods based on simple linear optimizations are proposed to generate wavelet filter banks with excellent frequency selectivity by maximizing the energy content in the filter passband. Finally, designed examples are given.
{"title":"A method of designing optimal wavelet filter banks","authors":"Peng Zhiwei, Wang Bo, Liao Guisheng","doi":"10.1109/ICOSP.1998.770200","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770200","url":null,"abstract":"Compactly supported orthogonal wavelets are obtained from two-band paraunitary FIR filter bank solutions, with the additional \"flatness\" constraint that the low-pass filter should have K zeros at half of the sampling frequency. This constraint is set to obtain a \"regular\" wavelet, but it is somewhat in contradiction with the requirement of excellent frequency selectivity. Maximally flat filters yielding Daubechies wavelets have the worst frequency selectivity. According to those drawbacks, two design methods based on simple linear optimizations are proposed to generate wavelet filter banks with excellent frequency selectivity by maximizing the energy content in the filter passband. Finally, designed examples are given.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"52 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":"131018750","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.1998.770288
Lixin Bao, J. Toyama, M. Shimbo
This paper proposes mixture hidden Markov models (HMM) with a time duration function to solve the recognition of Mandarin Chinese diphthongs and several words that resemble diphthongs. We propose an autoregression model to represent the dynamical relationships of observation symbols with time variance. The model can improve the weaknesses of standard HMM and nonstationary HMM.
{"title":"Mandarin phonetic recognition using mixture hidden Markov models with time duration function","authors":"Lixin Bao, J. Toyama, M. Shimbo","doi":"10.1109/ICOSP.1998.770288","DOIUrl":"https://doi.org/10.1109/ICOSP.1998.770288","url":null,"abstract":"This paper proposes mixture hidden Markov models (HMM) with a time duration function to solve the recognition of Mandarin Chinese diphthongs and several words that resemble diphthongs. We propose an autoregression model to represent the dynamical relationships of observation symbols with time variance. The model can improve the weaknesses of standard HMM and nonstationary HMM.","PeriodicalId":145700,"journal":{"name":"ICSP '98. 1998 Fourth International Conference on Signal Processing (Cat. No.98TH8344)","volume":"18 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":"130459497","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：