Pub Date : 2005-12-21DOI: 10.1109/ISSPIT.2005.1577075
A. Bartzas, G. Pouiklis, S. Mamagkakis, F. Catthoor, D. Soudris, A. Thanailakis
Demanding applications are introduced to networking field, asking simultaneously for high performance and low-energy consumption, requests more imperative in wireless networks. The dynamic nature of such applications makes the dynamic memory subsystem of an embedded system a critical contributing factor to the overall energy and execution time performance. This paper presents a novel aspect in designing dynamic data types in network applications. A systematic methodology, which is tool-supported and capable of manipulating different network traces, is proposed. Plethora of possible solutions, i.e. Pareto points, in terms of energy consumption, performance and memory size usage is achieved. Eventually, alternative optimal implementations, i.e. Pareto-optimal points can be extracted. Two real-life case studies (from NetBench suite) are studied and explored thoroughly. It is proved that up to 80% energy savings and up to 20% performance, comparing with two benchmarks' original implementation, can be accomplished. Furthermore, a plethora of trade-offs among the Pareto-optimal choices reach up to 52% for energy consumption and up to 13% for performance, are achieved
{"title":"Performance-energy trade-off exploration in dynamic data types for network applications","authors":"A. Bartzas, G. Pouiklis, S. Mamagkakis, F. Catthoor, D. Soudris, A. Thanailakis","doi":"10.1109/ISSPIT.2005.1577075","DOIUrl":"https://doi.org/10.1109/ISSPIT.2005.1577075","url":null,"abstract":"Demanding applications are introduced to networking field, asking simultaneously for high performance and low-energy consumption, requests more imperative in wireless networks. The dynamic nature of such applications makes the dynamic memory subsystem of an embedded system a critical contributing factor to the overall energy and execution time performance. This paper presents a novel aspect in designing dynamic data types in network applications. A systematic methodology, which is tool-supported and capable of manipulating different network traces, is proposed. Plethora of possible solutions, i.e. Pareto points, in terms of energy consumption, performance and memory size usage is achieved. Eventually, alternative optimal implementations, i.e. Pareto-optimal points can be extracted. Two real-life case studies (from NetBench suite) are studied and explored thoroughly. It is proved that up to 80% energy savings and up to 20% performance, comparing with two benchmarks' original implementation, can be accomplished. Furthermore, a plethora of trade-offs among the Pareto-optimal choices reach up to 52% for energy consumption and up to 13% for performance, are achieved","PeriodicalId":421826,"journal":{"name":"Proceedings of the Fifth IEEE International Symposium on Signal Processing and Information Technology, 2005.","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130292645","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 : 2005-12-21DOI: 10.1109/ISSPIT.2005.1577083
W. Cheng, Hu Wei-dong, Du Xiao-yong, Yu Wen-xian
In this paper, the authors present a new method of HRR profile formation based on the LFM signal fusion of the multiple radars with multiple frequency bands. The principle of the multiple radars signal fusion improving the range resolution is analyzed. During the fusion process, the B-splines interpolation and an entropy-minimization principle based phase compensation algorithm are applied. The theoretical analysis and simulations results show the proposed method can effectively increase signal bandwidth and provide a high resolution range profile
{"title":"A new method of HRR profile formation based on multiple radars LFM signal fusion","authors":"W. Cheng, Hu Wei-dong, Du Xiao-yong, Yu Wen-xian","doi":"10.1109/ISSPIT.2005.1577083","DOIUrl":"https://doi.org/10.1109/ISSPIT.2005.1577083","url":null,"abstract":"In this paper, the authors present a new method of HRR profile formation based on the LFM signal fusion of the multiple radars with multiple frequency bands. The principle of the multiple radars signal fusion improving the range resolution is analyzed. During the fusion process, the B-splines interpolation and an entropy-minimization principle based phase compensation algorithm are applied. The theoretical analysis and simulations results show the proposed method can effectively increase signal bandwidth and provide a high resolution range profile","PeriodicalId":421826,"journal":{"name":"Proceedings of the Fifth IEEE International Symposium on Signal Processing and Information Technology, 2005.","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117184587","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 : 2005-12-21DOI: 10.1109/ISSPIT.2005.1577217
N. Kroupis, N. Zervas, M. Dasygenis, K. Tatas, D. Soudris, A. Thanailakis
The continuous increase of the computational power of programmable processors has established them as an attractive design alternative, for implementation of the most computationally intensive applications, like video compression. To enforce this trend, designers implementing applications on programmable platforms have to be provided with reliable and in-depth analysis data that will allow for the early selection of the most appropriate application for a given set of specifications. To address this need, we introduce a new methodology for early and accurate estimation of the number of instructions required for the execution of an application, together with the number of data memory transfers on a programmable processor. The high-level estimation is achieved by a series of mathematical formulas; these describe not only the arithmetic operations of an application, but also its control and addressing operations, if it is executed on a programmable core. The comparative study, which is done using three popular processors (Pentium, ARM and MIPS), shows the high efficiency and accuracy of the methodology proposed, in terms of the number of executed (micro-)instructions (i.e. performance) and the number of data memory transfers (i.e. memory energy consumption)
{"title":"High-level performance and power exploration of DSP applications realized on programmable processors","authors":"N. Kroupis, N. Zervas, M. Dasygenis, K. Tatas, D. Soudris, A. Thanailakis","doi":"10.1109/ISSPIT.2005.1577217","DOIUrl":"https://doi.org/10.1109/ISSPIT.2005.1577217","url":null,"abstract":"The continuous increase of the computational power of programmable processors has established them as an attractive design alternative, for implementation of the most computationally intensive applications, like video compression. To enforce this trend, designers implementing applications on programmable platforms have to be provided with reliable and in-depth analysis data that will allow for the early selection of the most appropriate application for a given set of specifications. To address this need, we introduce a new methodology for early and accurate estimation of the number of instructions required for the execution of an application, together with the number of data memory transfers on a programmable processor. The high-level estimation is achieved by a series of mathematical formulas; these describe not only the arithmetic operations of an application, but also its control and addressing operations, if it is executed on a programmable core. The comparative study, which is done using three popular processors (Pentium, ARM and MIPS), shows the high efficiency and accuracy of the methodology proposed, in terms of the number of executed (micro-)instructions (i.e. performance) and the number of data memory transfers (i.e. memory energy consumption)","PeriodicalId":421826,"journal":{"name":"Proceedings of the Fifth IEEE International Symposium on Signal Processing and Information Technology, 2005.","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117232059","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 : 2005-12-21DOI: 10.1109/ISSPIT.2005.1577200
N. Eltonsy, E. Essock-Burns, G. Tourrasi, Adel Said Elmaghraby
This study analyzes the performance of a computer aided detection (CAD) scheme for mass detection in mammography. We investigate the trained parameters of the detection scheme before any further testing. We use an extended version of a previously reported mass detection scheme. We analyze the detection parameters by using linear canonical discriminants (LCD) and compare results with logistic regression and multi layer perceptron neural network models. Preliminary results suggest that regression and multi layer perceptron neural network showed the best receiver operator characteristics (ROC). The LCD analysis predictive function showed that the trained CAD scheme performance can maintain 99.08% sensitivity (108/109) with false positive rate (FPI) of 8 per image with ROC Az= 0.74plusmn0.01. The regression and the multi layer perceptron neural network ROC analysis showed stronger backbone for the CAD algorithm viewing that the extended CAD scheme can operate at 96% sensitivity with 5.6 FPI per image. These preliminary results suggest that further logic to reduce FPI is needed for the CAD algorithm to be more predictive
{"title":"Error investigation of models for improved detection of masses in screening mammography","authors":"N. Eltonsy, E. Essock-Burns, G. Tourrasi, Adel Said Elmaghraby","doi":"10.1109/ISSPIT.2005.1577200","DOIUrl":"https://doi.org/10.1109/ISSPIT.2005.1577200","url":null,"abstract":"This study analyzes the performance of a computer aided detection (CAD) scheme for mass detection in mammography. We investigate the trained parameters of the detection scheme before any further testing. We use an extended version of a previously reported mass detection scheme. We analyze the detection parameters by using linear canonical discriminants (LCD) and compare results with logistic regression and multi layer perceptron neural network models. Preliminary results suggest that regression and multi layer perceptron neural network showed the best receiver operator characteristics (ROC). The LCD analysis predictive function showed that the trained CAD scheme performance can maintain 99.08% sensitivity (108/109) with false positive rate (FPI) of 8 per image with ROC Az= 0.74plusmn0.01. The regression and the multi layer perceptron neural network ROC analysis showed stronger backbone for the CAD algorithm viewing that the extended CAD scheme can operate at 96% sensitivity with 5.6 FPI per image. These preliminary results suggest that further logic to reduce FPI is needed for the CAD algorithm to be more predictive","PeriodicalId":421826,"journal":{"name":"Proceedings of the Fifth IEEE International Symposium on Signal Processing and Information Technology, 2005.","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115846724","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 : 2005-12-21DOI: 10.1109/ISSPIT.2005.1577082
N. Petrochilos, G. Galati, L. Mene, E. Piracci
Multilateration systems based on secondary surveillance radar (SSR) and omni-directional antennae are operational today (P. Bezousek, 1998) with expected capacity limitations due to the increase of the air traffic. Assuming the replacement of the omni-directional antenna by an array, we proposed new algorithms in previous works (N. Petrochilos, July 2002), (N. Petrochilos, et al., May 2004). Unfortunately, they have some shortcomings, an expensive computational cost, and a not-so simple practical implementation. Therefore, there is a need for reliable, simple, effective algorithms such as the one we present here to separate multiple SSR signals. Real signals (as recorded in a live environment) are used to demonstrate the effectiveness of the proposed algorithm
{"title":"Separation of multiple secondary surveillance radar sources in a real environment by a novel projection algorithm","authors":"N. Petrochilos, G. Galati, L. Mene, E. Piracci","doi":"10.1109/ISSPIT.2005.1577082","DOIUrl":"https://doi.org/10.1109/ISSPIT.2005.1577082","url":null,"abstract":"Multilateration systems based on secondary surveillance radar (SSR) and omni-directional antennae are operational today (P. Bezousek, 1998) with expected capacity limitations due to the increase of the air traffic. Assuming the replacement of the omni-directional antenna by an array, we proposed new algorithms in previous works (N. Petrochilos, July 2002), (N. Petrochilos, et al., May 2004). Unfortunately, they have some shortcomings, an expensive computational cost, and a not-so simple practical implementation. Therefore, there is a need for reliable, simple, effective algorithms such as the one we present here to separate multiple SSR signals. Real signals (as recorded in a live environment) are used to demonstrate the effectiveness of the proposed algorithm","PeriodicalId":421826,"journal":{"name":"Proceedings of the Fifth IEEE International Symposium on Signal Processing and Information Technology, 2005.","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132473344","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 : 2005-12-21DOI: 10.1109/ISSPIT.2005.1577177
A. Mohamed, S. Rajasekaran
Text summarization is an important problem since it has numerous applications. This problem has been extensively studied and many approaches have been proposed in the literature for its solution. In this paper, we investigate a new approach that employs meta-search. In particular, summaries from several summarizers are evaluated to identify the best summary. To the best of our knowledge, this is the first work that employs meta-search in the context of text summarization. We have employed data from Document Understanding Conference 2002 (DUC-2002) and 5 different summarizers in our experiments
{"title":"A text summarizer based on meta-search","authors":"A. Mohamed, S. Rajasekaran","doi":"10.1109/ISSPIT.2005.1577177","DOIUrl":"https://doi.org/10.1109/ISSPIT.2005.1577177","url":null,"abstract":"Text summarization is an important problem since it has numerous applications. This problem has been extensively studied and many approaches have been proposed in the literature for its solution. In this paper, we investigate a new approach that employs meta-search. In particular, summaries from several summarizers are evaluated to identify the best summary. To the best of our knowledge, this is the first work that employs meta-search in the context of text summarization. We have employed data from Document Understanding Conference 2002 (DUC-2002) and 5 different summarizers in our experiments","PeriodicalId":421826,"journal":{"name":"Proceedings of the Fifth IEEE International Symposium on Signal Processing and Information Technology, 2005.","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114419020","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 : 2005-12-21DOI: 10.1109/ISSPIT.2005.1577073
A. Hussein, R. Ammar, A. Hamdy
Scheduling real-time applications requires high utilization of the available processing power on processors to accommodate as many tasks as possible while satisfying the required deadlines of the applications. Moreover, an acceptable level of reliability should be provided due to the critical nature of the applications. In this paper, we propose a new approach that highly utilizes the available processing power to improve the admission rate of real-time task scheduling. The basic idea is to use all of the available processing power to execute allocated tasks. This reduces the execution time needed for each task and hence increases its reliability. It also reduces needed processing power for existing tasks to satisfy their deadlines and hence provides higher available processing power for any new task. This gives a higher chance for the new arriving task to be accommodated. We show that our approach is beneficial in that it provides a better scheme with higher acceptance rate for real time tasks compared to the traditional approach of using a fixed processing power reservation
{"title":"Efficient processing power utilization to execute real-time tasks","authors":"A. Hussein, R. Ammar, A. Hamdy","doi":"10.1109/ISSPIT.2005.1577073","DOIUrl":"https://doi.org/10.1109/ISSPIT.2005.1577073","url":null,"abstract":"Scheduling real-time applications requires high utilization of the available processing power on processors to accommodate as many tasks as possible while satisfying the required deadlines of the applications. Moreover, an acceptable level of reliability should be provided due to the critical nature of the applications. In this paper, we propose a new approach that highly utilizes the available processing power to improve the admission rate of real-time task scheduling. The basic idea is to use all of the available processing power to execute allocated tasks. This reduces the execution time needed for each task and hence increases its reliability. It also reduces needed processing power for existing tasks to satisfy their deadlines and hence provides higher available processing power for any new task. This gives a higher chance for the new arriving task to be accommodated. We show that our approach is beneficial in that it provides a better scheme with higher acceptance rate for real time tasks compared to the traditional approach of using a fixed processing power reservation","PeriodicalId":421826,"journal":{"name":"Proceedings of the Fifth IEEE International Symposium on Signal Processing and Information Technology, 2005.","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121996350","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 : 2005-12-21DOI: 10.1109/ISSPIT.2005.1577213
F. Zargari, O. Fatemi
It is well known that images of natural scenes have predominantly low frequency components. Wavelet transform (WT) which is used in JPEG 2000, concentrates the low frequency of the transformed image in LL sub-band. Hence, transmission error in LL sub-band of JPEG 2000 coded image affects the quality of the perceived image drastically. Another important aspect of JPEG 2000 standard is employing large code-blocks (normally 32times32 or larger) to provide sufficient compression efficiency. Error concealment (EC) in so large areas is a challenging problem. Filling the lost bits with zero (ZF) is the only well known EC method for large areas. In this paper the segmentation and conditional replacement (SCR) EC method is introduced for EC of LL sub-band. This method is based on the segmentation and replacement (SR) EC method and outperforms the ZF method at large areas for any number of lost bit planes in LL sub-band
{"title":"Recovery of damaged code blocks in LL sub-band of JPEG 2000 images","authors":"F. Zargari, O. Fatemi","doi":"10.1109/ISSPIT.2005.1577213","DOIUrl":"https://doi.org/10.1109/ISSPIT.2005.1577213","url":null,"abstract":"It is well known that images of natural scenes have predominantly low frequency components. Wavelet transform (WT) which is used in JPEG 2000, concentrates the low frequency of the transformed image in LL sub-band. Hence, transmission error in LL sub-band of JPEG 2000 coded image affects the quality of the perceived image drastically. Another important aspect of JPEG 2000 standard is employing large code-blocks (normally 32times32 or larger) to provide sufficient compression efficiency. Error concealment (EC) in so large areas is a challenging problem. Filling the lost bits with zero (ZF) is the only well known EC method for large areas. In this paper the segmentation and conditional replacement (SCR) EC method is introduced for EC of LL sub-band. This method is based on the segmentation and replacement (SR) EC method and outperforms the ZF method at large areas for any number of lost bit planes in LL sub-band","PeriodicalId":421826,"journal":{"name":"Proceedings of the Fifth IEEE International Symposium on Signal Processing and Information Technology, 2005.","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129857440","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 : 2005-12-21DOI: 10.1109/ISSPIT.2005.1577097
D. Rouvre, D. Kouamé, F. Tranquart, L. Pourcelot
This paper presents a new technique called empirical mode decomposition (EMD) applied to a multi-gate, multitransducer ultrasound Doppler system used for fetal heart monitoring. We propose this system as an alternative to the existing fetal monitoring techniques. Classical autocorrelation-based fetal heart rate (FHR) detection has been shown to be a good method to detect the FHR in normal situations. However, as this method is applied to magnitudes of the Doppler signal fails to estimate the fetal heart rate when the fetus moves. In view of the extent of FHR variability, a monitoring system should be able to estimate this parameter each time. We therefore propose empirical mode decomposition as an interesting alternative for long-term monitoring. The principle of this method consists of iterative decompositions of a signal into a sum of functions that have the same number of extrema, the same number of zero crossings, and are symmetric with respect to the local mean. When investigation of FHR using autocorrelation on the original Doppler signal fails due to fetal movement or low signal-to-noise ratio (SNR), it is frequently successful using the intrinsic mode functions (imfs). We compared the results of multi-transducer FHR detection with and without EMD decomposition using in-vivo Doppler signals from a set of 40 women between 32 and 38 weeks of pregnancy. This method greatly improved the quality of FHR detection
{"title":"Empirical mode decomposition (EMD) for multi-gate, multi-transducer ultrasound Doppler fetal heart monitoring","authors":"D. Rouvre, D. Kouamé, F. Tranquart, L. Pourcelot","doi":"10.1109/ISSPIT.2005.1577097","DOIUrl":"https://doi.org/10.1109/ISSPIT.2005.1577097","url":null,"abstract":"This paper presents a new technique called empirical mode decomposition (EMD) applied to a multi-gate, multitransducer ultrasound Doppler system used for fetal heart monitoring. We propose this system as an alternative to the existing fetal monitoring techniques. Classical autocorrelation-based fetal heart rate (FHR) detection has been shown to be a good method to detect the FHR in normal situations. However, as this method is applied to magnitudes of the Doppler signal fails to estimate the fetal heart rate when the fetus moves. In view of the extent of FHR variability, a monitoring system should be able to estimate this parameter each time. We therefore propose empirical mode decomposition as an interesting alternative for long-term monitoring. The principle of this method consists of iterative decompositions of a signal into a sum of functions that have the same number of extrema, the same number of zero crossings, and are symmetric with respect to the local mean. When investigation of FHR using autocorrelation on the original Doppler signal fails due to fetal movement or low signal-to-noise ratio (SNR), it is frequently successful using the intrinsic mode functions (imfs). We compared the results of multi-transducer FHR detection with and without EMD decomposition using in-vivo Doppler signals from a set of 40 women between 32 and 38 weeks of pregnancy. This method greatly improved the quality of FHR detection","PeriodicalId":421826,"journal":{"name":"Proceedings of the Fifth IEEE International Symposium on Signal Processing and Information Technology, 2005.","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129964070","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 : 2005-12-21DOI: 10.1109/ISSPIT.2005.1577163
J. Gllavata, Bernd Freisleben
The extraction of textual information from images and videos is an important task for automatic content-based indexing and retrieval purposes. To extract text from images or videos coming from unknown international sources, it is necessary to know the script beforehand in order to employ suitable text segmentation and optical character recognition (OCR) methods. In this paper, we present an approach for discriminating between Latin and Ideographic script. The proposed approach proceeds as follows: first, the text present in an image is localized. Then, a set of low-level features is extracted from the localized text image. Finally, based on the extracted features, the decision about the type of the script is made using a k-nearest neighbour classifier. Initial experimental results for a set of images containing text of different scripts demonstrate the good performance of the proposed solution
{"title":"Script recognition in images with complex backgrounds","authors":"J. Gllavata, Bernd Freisleben","doi":"10.1109/ISSPIT.2005.1577163","DOIUrl":"https://doi.org/10.1109/ISSPIT.2005.1577163","url":null,"abstract":"The extraction of textual information from images and videos is an important task for automatic content-based indexing and retrieval purposes. To extract text from images or videos coming from unknown international sources, it is necessary to know the script beforehand in order to employ suitable text segmentation and optical character recognition (OCR) methods. In this paper, we present an approach for discriminating between Latin and Ideographic script. The proposed approach proceeds as follows: first, the text present in an image is localized. Then, a set of low-level features is extracted from the localized text image. Finally, based on the extracted features, the decision about the type of the script is made using a k-nearest neighbour classifier. Initial experimental results for a set of images containing text of different scripts demonstrate the good performance of the proposed solution","PeriodicalId":421826,"journal":{"name":"Proceedings of the Fifth IEEE International Symposium on Signal Processing and Information Technology, 2005.","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121669020","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：