Pub Date : 1998-09-07DOI: 10.1109/COMSIG.1998.737010
A. Wilkinson, M. R. Inggs
The development of new sensors for landmine detection is an active area of research. Microwave radiometry is a field of remote sensing concerned with the measurement of thermally emitted radiation in the microwave band. This paper describes the design, testing and performance of an experimental microwave radiometric measurement system for use in landmine detection. The aim of our research is to set up an experiment to assess the potential of microwave radiometry for landmine detection and also to establish the relative merits of various frequency bands. We have designed and constructed a sensitive radiometric receiver consisting of a switchable front end (i.e., 10 GHz, 35 GHz or 90 GHz), followed by an IF amplifier chain of bandwidth 1 GHz, and power detector. The radiometer design, hardware and performance are described.
{"title":"Radiometry for landmine detection","authors":"A. Wilkinson, M. R. Inggs","doi":"10.1109/COMSIG.1998.737010","DOIUrl":"https://doi.org/10.1109/COMSIG.1998.737010","url":null,"abstract":"The development of new sensors for landmine detection is an active area of research. Microwave radiometry is a field of remote sensing concerned with the measurement of thermally emitted radiation in the microwave band. This paper describes the design, testing and performance of an experimental microwave radiometric measurement system for use in landmine detection. The aim of our research is to set up an experiment to assess the potential of microwave radiometry for landmine detection and also to establish the relative merits of various frequency bands. We have designed and constructed a sensitive radiometric receiver consisting of a switchable front end (i.e., 10 GHz, 35 GHz or 90 GHz), followed by an IF amplifier chain of bandwidth 1 GHz, and power detector. The radiometer design, hardware and performance are described.","PeriodicalId":294473,"journal":{"name":"Proceedings of the 1998 South African Symposium on Communications and Signal Processing-COMSIG '98 (Cat. No. 98EX214)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115495259","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 : 1998-09-07DOI: 10.1109/COMSIG.1998.736948
J. du Preez, D. Weber
Recently we introduced algorithms for the efficient processing of high-order hidden Markov models (HMMs) of fixed order into the literature. In this paper these techniques are generalised to also include mixed-order HMMs. This then allows first-order equivalents to be found for all HMMs, thereby providing a unifying base for reasoning about their properties. Specifically we provide formulations that separates the duration modelling and context modelling capabilities of high-order HMMs.
{"title":"High-order hidden Markov modelling","authors":"J. du Preez, D. Weber","doi":"10.1109/COMSIG.1998.736948","DOIUrl":"https://doi.org/10.1109/COMSIG.1998.736948","url":null,"abstract":"Recently we introduced algorithms for the efficient processing of high-order hidden Markov models (HMMs) of fixed order into the literature. In this paper these techniques are generalised to also include mixed-order HMMs. This then allows first-order equivalents to be found for all HMMs, thereby providing a unifying base for reasoning about their properties. Specifically we provide formulations that separates the duration modelling and context modelling capabilities of high-order HMMs.","PeriodicalId":294473,"journal":{"name":"Proceedings of the 1998 South African Symposium on Communications and Signal Processing-COMSIG '98 (Cat. No. 98EX214)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129870960","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 : 1998-09-07DOI: 10.1109/COMSIG.1998.736994
R. Hansmann, D. Davidson
Edge-based vector finite elements have been implemented for 2D eigenvalue problems. The formulation of the hybrid FEM-BEM method for scattering problems is described and application to simple scattering configurations is discussed.
{"title":"Two-dimensional edge-based finite elements for guided and scattered wave problems","authors":"R. Hansmann, D. Davidson","doi":"10.1109/COMSIG.1998.736994","DOIUrl":"https://doi.org/10.1109/COMSIG.1998.736994","url":null,"abstract":"Edge-based vector finite elements have been implemented for 2D eigenvalue problems. The formulation of the hybrid FEM-BEM method for scattering problems is described and application to simple scattering configurations is discussed.","PeriodicalId":294473,"journal":{"name":"Proceedings of the 1998 South African Symposium on Communications and Signal Processing-COMSIG '98 (Cat. No. 98EX214)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121874544","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 : 1998-09-07DOI: 10.1109/COMSIG.1998.736924
K. Scheffler, J. du Preez
The field of topic spotting in conversational speech has been receiving growing attention in recent years. The goal of this field is to develop a system that can identify topics of interest among large volumes of speech data. In order to cope with practical considerations, researchers are concentrating on phoneme-based methods, which eliminate the need for topic specific data to be hand-transcribed. A number of different phoneme-based approaches have recently been proposed, of which the Euclidean nearest wrong neighbour (ENWN) system (Kuhn et al, 1997) has yielded the most promising experimental results. A phoneme-based topic spotter makes use of a phoneme recogniser to transcribe the speech data. The main problem of this approach is that the accuracy of such transcriptions is very poor. Typically, only between 40 and 50 percent of the phonemes are transcribed correctly. It is therefore important to compensate for the low quality of the transcriptions. However, existing techniques make no use of statistical modelling to compensate for transcription errors. In this research, a stochastic method for automatic recognition of topics (SMART) was developed to address the above mentioned problem. The resulting system is an extension of the existing ENWN algorithm. Comparative results indicate an improvement of SMART over ENWN characterized by a 26% reduction in ROC (receiver operating characteristic) error area. This difference was found to be statistically significant.
近年来，会话话语中的话题定位受到越来越多的关注。该领域的目标是开发一个能够在大量语音数据中识别感兴趣主题的系统。为了应对实际的考虑，研究人员正在集中研究基于音素的方法，这种方法消除了对特定主题数据的手工转录的需要。最近提出了许多不同的基于音素的方法，其中欧几里得最近邻错误(ENWN)系统(Kuhn et al, 1997)产生了最有希望的实验结果。基于音素的主题识别器利用音素识别器转录语音数据。这种方法的主要问题是这种转录的准确性很差。通常情况下，只有40%到50%的音素被正确转录。因此，重要的是要弥补转录的低质量。然而，现有的技术没有使用统计模型来补偿转录错误。为了解决上述问题，本研究提出了一种随机主题自动识别方法(SMART)。所得到的系统是现有ENWN算法的扩展。对比结果表明，与ENWN相比，SMART的改进表现为ROC(接受者工作特征)误差面积减少了26%。这种差异在统计学上是显著的。
{"title":"Stochastic method for automatic recognition of topics","authors":"K. Scheffler, J. du Preez","doi":"10.1109/COMSIG.1998.736924","DOIUrl":"https://doi.org/10.1109/COMSIG.1998.736924","url":null,"abstract":"The field of topic spotting in conversational speech has been receiving growing attention in recent years. The goal of this field is to develop a system that can identify topics of interest among large volumes of speech data. In order to cope with practical considerations, researchers are concentrating on phoneme-based methods, which eliminate the need for topic specific data to be hand-transcribed. A number of different phoneme-based approaches have recently been proposed, of which the Euclidean nearest wrong neighbour (ENWN) system (Kuhn et al, 1997) has yielded the most promising experimental results. A phoneme-based topic spotter makes use of a phoneme recogniser to transcribe the speech data. The main problem of this approach is that the accuracy of such transcriptions is very poor. Typically, only between 40 and 50 percent of the phonemes are transcribed correctly. It is therefore important to compensate for the low quality of the transcriptions. However, existing techniques make no use of statistical modelling to compensate for transcription errors. In this research, a stochastic method for automatic recognition of topics (SMART) was developed to address the above mentioned problem. The resulting system is an extension of the existing ENWN algorithm. Comparative results indicate an improvement of SMART over ENWN characterized by a 26% reduction in ROC (receiver operating characteristic) error area. This difference was found to be statistically significant.","PeriodicalId":294473,"journal":{"name":"Proceedings of the 1998 South African Symposium on Communications and Signal Processing-COMSIG '98 (Cat. No. 98EX214)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122171332","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 : 1998-09-07DOI: 10.1109/COMSIG.1998.736949
H. Serretta, M. Inggs
This paper summarises further research into the application of the Mellin transform for radar ship target recognition. As reported in previous papers, preprocessing is via the Mellin transform, with recognition utilising neural networks. However, in this paper, substantial emphasis is placed on discussing the pre-processing techniques applied in the implementation of the Fourier modified direct Mellin transform (FMDMT). The FDMDMT extracts features from the range profiles of targets, regardless of aspect angle. In particular, the robustness of the technique in the presence of noise and amplitude scintillation is investigated, together with the range of angles over which the FMDMT is successful. Real and simulated data is utilised.
{"title":"Ship target recognition with the Mellin transform aided by neural networks","authors":"H. Serretta, M. Inggs","doi":"10.1109/COMSIG.1998.736949","DOIUrl":"https://doi.org/10.1109/COMSIG.1998.736949","url":null,"abstract":"This paper summarises further research into the application of the Mellin transform for radar ship target recognition. As reported in previous papers, preprocessing is via the Mellin transform, with recognition utilising neural networks. However, in this paper, substantial emphasis is placed on discussing the pre-processing techniques applied in the implementation of the Fourier modified direct Mellin transform (FMDMT). The FDMDMT extracts features from the range profiles of targets, regardless of aspect angle. In particular, the robustness of the technique in the presence of noise and amplitude scintillation is investigated, together with the range of angles over which the FMDMT is successful. Real and simulated data is utilised.","PeriodicalId":294473,"journal":{"name":"Proceedings of the 1998 South African Symposium on Communications and Signal Processing-COMSIG '98 (Cat. No. 98EX214)","volume":"164 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123159818","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 : 1998-09-07DOI: 10.1109/COMSIG.1998.736934
Dani Kramer, F. Aghdasi
We present a multiscale statistical approach to texture analysis. These techniques are used to classify microcalcifications in digitised mammograms as benign or malignant. In this study we extract the proposed multiscale statistical texture signatures, based on the co-occurrence matrix, as well as wavelet-based texture signatures from the regions of interest containing the microcalcifications. The discriminatory ability of these texture signatures is demonstrated by their ability to successfully distinguish between benign and malignant cases. Classification is performed by means of a k-nearest neighbour classifier. One hundred percent correct classification is achieved when using a combination of the multiscale statistical texture signatures and the wavelet-based texture signatures. A database with a small number of samples was used, and further analysis with a larger database will give these results greater statistical significance.
{"title":"Classification of microcalcifications in digitised mammograms using multiscale statistical texture analysis","authors":"Dani Kramer, F. Aghdasi","doi":"10.1109/COMSIG.1998.736934","DOIUrl":"https://doi.org/10.1109/COMSIG.1998.736934","url":null,"abstract":"We present a multiscale statistical approach to texture analysis. These techniques are used to classify microcalcifications in digitised mammograms as benign or malignant. In this study we extract the proposed multiscale statistical texture signatures, based on the co-occurrence matrix, as well as wavelet-based texture signatures from the regions of interest containing the microcalcifications. The discriminatory ability of these texture signatures is demonstrated by their ability to successfully distinguish between benign and malignant cases. Classification is performed by means of a k-nearest neighbour classifier. One hundred percent correct classification is achieved when using a combination of the multiscale statistical texture signatures and the wavelet-based texture signatures. A database with a small number of samples was used, and further analysis with a larger database will give these results greater statistical significance.","PeriodicalId":294473,"journal":{"name":"Proceedings of the 1998 South African Symposium on Communications and Signal Processing-COMSIG '98 (Cat. No. 98EX214)","volume":"89 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125670212","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 : 1998-09-07DOI: 10.1109/COMSIG.1998.736916
G. Venayagamoorthy, V. Moonasar, K. Sandrasegaran
One solution to the crime and illegal immigration problem in South Africa may be the use of biometric techniques and technology. Biometrics are methods for recognizing a user based on unique physiological and/or behavioural characteristics of the user. This paper presents the results of ongoing work into using neural networks for voice recognition.
{"title":"Voice recognition using neural networks","authors":"G. Venayagamoorthy, V. Moonasar, K. Sandrasegaran","doi":"10.1109/COMSIG.1998.736916","DOIUrl":"https://doi.org/10.1109/COMSIG.1998.736916","url":null,"abstract":"One solution to the crime and illegal immigration problem in South Africa may be the use of biometric techniques and technology. Biometrics are methods for recognizing a user based on unique physiological and/or behavioural characteristics of the user. This paper presents the results of ongoing work into using neural networks for voice recognition.","PeriodicalId":294473,"journal":{"name":"Proceedings of the 1998 South African Symposium on Communications and Signal Processing-COMSIG '98 (Cat. No. 98EX214)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129231456","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 : 1998-09-07DOI: 10.1109/COMSIG.1998.736932
J. Horrell, M. Inggs
This paper investigates the effects of severe range curvature on range-Doppler SAR processing for systems with centre frequencies in the range of 100-200 MHz. The focus is on airborne, strip-map, side-looking SAR systems without range walk (i.e. imaging geometries with a Doppler centroid of zero). The effects on the focusing of a simulated point target are investigated. The main objective is to provide an indication to users of range-Doppler processors what performance could be expected at VHF frequencies when no Doppler-frequency-dependent secondary range compression (SRC) is performed.
{"title":"Low frequency range-Doppler SAR processing without secondary range compression","authors":"J. Horrell, M. Inggs","doi":"10.1109/COMSIG.1998.736932","DOIUrl":"https://doi.org/10.1109/COMSIG.1998.736932","url":null,"abstract":"This paper investigates the effects of severe range curvature on range-Doppler SAR processing for systems with centre frequencies in the range of 100-200 MHz. The focus is on airborne, strip-map, side-looking SAR systems without range walk (i.e. imaging geometries with a Doppler centroid of zero). The effects on the focusing of a simulated point target are investigated. The main objective is to provide an indication to users of range-Doppler processors what performance could be expected at VHF frequencies when no Doppler-frequency-dependent secondary range compression (SRC) is performed.","PeriodicalId":294473,"journal":{"name":"Proceedings of the 1998 South African Symposium on Communications and Signal Processing-COMSIG '98 (Cat. No. 98EX214)","volume":"12 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131957490","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 : 1998-09-07DOI: 10.1109/COMSIG.1998.736962
Ashok Sahai
The motivation behind this paper is threefold, as follows. Firstly, it attempts to make a comprehensive exposition of the important role of elementary 'information theory', relevant to 'communication coding'. Secondly, the paper tries to illustrate comprehensively and convincingly how 'information theory' is a bridge between two different disciplines, namely 'communication and coding' and 'probability and statistics'. Thirdly, and most importantly, the paper leads to the conclusion that while we are trying to solve a problem in a particular area, e.g. 'electrical engineering', we ought to have a 'research group' with scientists from all relevant areas, for instance economics, probability and statistics, geology etc. The point is: by marrying expertise in seemingly diverse fields of knowledge we would vastly improve our chances of arriving at a gainful solution to the problem at hand, which would be practically optimal.
{"title":"Communication and coding vis-a-vis information theory vis-a-vis probability and statistics","authors":"Ashok Sahai","doi":"10.1109/COMSIG.1998.736962","DOIUrl":"https://doi.org/10.1109/COMSIG.1998.736962","url":null,"abstract":"The motivation behind this paper is threefold, as follows. Firstly, it attempts to make a comprehensive exposition of the important role of elementary 'information theory', relevant to 'communication coding'. Secondly, the paper tries to illustrate comprehensively and convincingly how 'information theory' is a bridge between two different disciplines, namely 'communication and coding' and 'probability and statistics'. Thirdly, and most importantly, the paper leads to the conclusion that while we are trying to solve a problem in a particular area, e.g. 'electrical engineering', we ought to have a 'research group' with scientists from all relevant areas, for instance economics, probability and statistics, geology etc. The point is: by marrying expertise in seemingly diverse fields of knowledge we would vastly improve our chances of arriving at a gainful solution to the problem at hand, which would be practically optimal.","PeriodicalId":294473,"journal":{"name":"Proceedings of the 1998 South African Symposium on Communications and Signal Processing-COMSIG '98 (Cat. No. 98EX214)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125081515","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 : 1998-09-07DOI: 10.1109/COMSIG.1998.736953
A.H. Bredekamp, J. Tapson
This paper describes the novel design and implementation of a fast feedback controller for a scanning tunneling microscope (STM). The system, which is required to control the Z-axis displacement of the piezoceramic STM scanner, replaces the conventional DSP controller with fast 8-bit microcontrollers which have only limited DSP capability. The lateral or XY control movement is implemented on a separate microcontroller. The control electronics are connected to a PC, which performs the image storage, analysis and display.
{"title":"A scanning tunneling microscope control system based on fast microcontrollers","authors":"A.H. Bredekamp, J. Tapson","doi":"10.1109/COMSIG.1998.736953","DOIUrl":"https://doi.org/10.1109/COMSIG.1998.736953","url":null,"abstract":"This paper describes the novel design and implementation of a fast feedback controller for a scanning tunneling microscope (STM). The system, which is required to control the Z-axis displacement of the piezoceramic STM scanner, replaces the conventional DSP controller with fast 8-bit microcontrollers which have only limited DSP capability. The lateral or XY control movement is implemented on a separate microcontroller. The control electronics are connected to a PC, which performs the image storage, analysis and display.","PeriodicalId":294473,"journal":{"name":"Proceedings of the 1998 South African Symposium on Communications and Signal Processing-COMSIG '98 (Cat. No. 98EX214)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125261084","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：