Pub Date : 1992-08-30DOI: 10.1109/ICPR.1992.201803
A. Laurentini, P. Viada
Tables are important components of technical documents. This paper addresses the following problems: (i) identifying a tabular component in a scanned image of a compound document containing text, drawings, diagrams, etc.; (ii) understanding the content of the table in order to convert the table into electronic format. As far as the authors are aware, the problems addressed are new. An algorithm for performing both the above tasks has been studied and implemented. Preliminary experimental results indicate satisfactory performance for many table lay-out styles.<>
{"title":"Identifying and understanding tabular material in compound documents","authors":"A. Laurentini, P. Viada","doi":"10.1109/ICPR.1992.201803","DOIUrl":"https://doi.org/10.1109/ICPR.1992.201803","url":null,"abstract":"Tables are important components of technical documents. This paper addresses the following problems: (i) identifying a tabular component in a scanned image of a compound document containing text, drawings, diagrams, etc.; (ii) understanding the content of the table in order to convert the table into electronic format. As far as the authors are aware, the problems addressed are new. An algorithm for performing both the above tasks has been studied and implemented. Preliminary experimental results indicate satisfactory performance for many table lay-out styles.<<ETX>>","PeriodicalId":34917,"journal":{"name":"模式识别与人工智能","volume":"69 1","pages":"405-409"},"PeriodicalIF":0.0,"publicationDate":"1992-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85882822","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 : 1992-08-30DOI: 10.1109/ICPR.1992.201719
G. Tambouratzis, T. Stonham
An online, unsupervised training algorithm is presented, which allows a logical neural network already trained to identify classes of objects to adapt to changes in the environment. This algorithm enables the system to operate continuously, without danger of overgeneralisation and displays useful noise-reduction properties. Results indicating its capabilities and characteristics in this adaptation task are described. The algorithm's self-organisation characteristics are also evaluated.<>
{"title":"A logical neural network that adapts to changes in the pattern environment","authors":"G. Tambouratzis, T. Stonham","doi":"10.1109/ICPR.1992.201719","DOIUrl":"https://doi.org/10.1109/ICPR.1992.201719","url":null,"abstract":"An online, unsupervised training algorithm is presented, which allows a logical neural network already trained to identify classes of objects to adapt to changes in the environment. This algorithm enables the system to operate continuously, without danger of overgeneralisation and displays useful noise-reduction properties. Results indicating its capabilities and characteristics in this adaptation task are described. The algorithm's self-organisation characteristics are also evaluated.<<ETX>>","PeriodicalId":34917,"journal":{"name":"模式识别与人工智能","volume":"7 1","pages":"46-49"},"PeriodicalIF":0.0,"publicationDate":"1992-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87792753","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 : 1992-08-30DOI: 10.1109/ICPR.1992.202124
P. Jonker, J. J. Gerbrands
A real-time pipelined image processing system operating in a time division multiplexing mode to serve up to 16 cameras, was realized to count the mass of a flow of bottles on a conveyor belt. The realized mass counting system proved to be a powerful tool capable of continuously counting bottles with a speed of approximately 500000 bottles a day per measurement point and with an accuracy of less then 0.5%.<>
{"title":"Image processing hardware for counting massive object streams","authors":"P. Jonker, J. J. Gerbrands","doi":"10.1109/ICPR.1992.202124","DOIUrl":"https://doi.org/10.1109/ICPR.1992.202124","url":null,"abstract":"A real-time pipelined image processing system operating in a time division multiplexing mode to serve up to 16 cameras, was realized to count the mass of a flow of bottles on a conveyor belt. The realized mass counting system proved to be a powerful tool capable of continuously counting bottles with a speed of approximately 500000 bottles a day per measurement point and with an accuracy of less then 0.5%.<<ETX>>","PeriodicalId":34917,"journal":{"name":"模式识别与人工智能","volume":"124 1","pages":"31-33"},"PeriodicalIF":0.0,"publicationDate":"1992-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88086030","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 : 1992-08-30DOI: 10.1109/ICPR.1992.201862
Shuh-Chuan Tsay, Peir-Ren Hong, Bin-Chang Chieu
Performs the handwritten digits recognition using the OCON (one-class-one-net) network and the PSU (pruning selective update) training algorithm. The main feature of the architecture of OCON network is that the entire network is composed of single output multi-layer perceptron and each of the subnets represents one class. The PSU training algorithm defined on the new cost function is designed to speed up the training procedure. It is shown that an OCON network with the new training algorithm outperforms the conventional back-propagation algorithm.<>
{"title":"Handwritten digits recognition system via OCON neural network by pruning selective update","authors":"Shuh-Chuan Tsay, Peir-Ren Hong, Bin-Chang Chieu","doi":"10.1109/ICPR.1992.201862","DOIUrl":"https://doi.org/10.1109/ICPR.1992.201862","url":null,"abstract":"Performs the handwritten digits recognition using the OCON (one-class-one-net) network and the PSU (pruning selective update) training algorithm. The main feature of the architecture of OCON network is that the entire network is composed of single output multi-layer perceptron and each of the subnets represents one class. The PSU training algorithm defined on the new cost function is designed to speed up the training procedure. It is shown that an OCON network with the new training algorithm outperforms the conventional back-propagation algorithm.<<ETX>>","PeriodicalId":34917,"journal":{"name":"模式识别与人工智能","volume":"16 1","pages":"656-659"},"PeriodicalIF":0.0,"publicationDate":"1992-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86860880","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 : 1992-08-30DOI: 10.1109/ICPR.1992.201846
Liang Li, T. Ho, J. Hull, S. Srihari
A top-down approach to word recognition is proposed. Discussions are presented on dynamically selecting the most effective feature combinations, which are applied to discriminate between a limited set of word hypotheses.<>
提出了一种自顶向下的词识别方法。讨论了动态选择最有效的特征组合，用于区分有限的一组词假设。
{"title":"A hypothesis testing approach to word recognition using dynamic feature selection","authors":"Liang Li, T. Ho, J. Hull, S. Srihari","doi":"10.1109/ICPR.1992.201846","DOIUrl":"https://doi.org/10.1109/ICPR.1992.201846","url":null,"abstract":"A top-down approach to word recognition is proposed. Discussions are presented on dynamically selecting the most effective feature combinations, which are applied to discriminate between a limited set of word hypotheses.<<ETX>>","PeriodicalId":34917,"journal":{"name":"模式识别与人工智能","volume":"35 1","pages":"586-589"},"PeriodicalIF":0.0,"publicationDate":"1992-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87112042","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 : 1992-08-30DOI: 10.1109/ICPR.1992.201852
L. Cordella, C. Stefano, F. Tortorella, M. Vento
A neural classifier for isolated omnifont characters is discussed. A method for characterizing a given training set of characters, based on the definition of some statistical parameters is introduced; on the basis of such characterization an architecture is defined made of a set of neural networks properly connected. Depending on the value of the parameters characterizing the training set, both sizing and training of each network are separately carried out according to a suitable methodology. It is shown that higher recognition rates can be achieved than those obtained by using a single neural network as classifier.<>
{"title":"Improving character recognition rate by a multi-net neural classifier","authors":"L. Cordella, C. Stefano, F. Tortorella, M. Vento","doi":"10.1109/ICPR.1992.201852","DOIUrl":"https://doi.org/10.1109/ICPR.1992.201852","url":null,"abstract":"A neural classifier for isolated omnifont characters is discussed. A method for characterizing a given training set of characters, based on the definition of some statistical parameters is introduced; on the basis of such characterization an architecture is defined made of a set of neural networks properly connected. Depending on the value of the parameters characterizing the training set, both sizing and training of each network are separately carried out according to a suitable methodology. It is shown that higher recognition rates can be achieved than those obtained by using a single neural network as classifier.<<ETX>>","PeriodicalId":34917,"journal":{"name":"模式识别与人工智能","volume":"75 1","pages":"615-618"},"PeriodicalIF":0.0,"publicationDate":"1992-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86259647","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 : 1992-08-30DOI: 10.1109/ICPR.1992.201770
J. Gloger
Presents a new method for the separation of merged text/form-structure components in forms. The technique described uses a modified version of the Hough transform to detect the structure of the form. The closed contours of the connected components are approximated by piecewise linear line segments. The parameters of the Hesse normal form of each line segment serve as input for the Hough transform. Compared to the vectorized boundary of characters, the lines of the form structure consist of appreciable more line segments with the same orientation and distance. So, the problem of the form structure detection in the database of line segments can be reduced to the detection of local peaks in the Hough space. Subsequent processing steps reconstruct the remaining contour fragments to characters.<>
{"title":"Use of the Hough transform to separate merged text/graphics in forms","authors":"J. Gloger","doi":"10.1109/ICPR.1992.201770","DOIUrl":"https://doi.org/10.1109/ICPR.1992.201770","url":null,"abstract":"Presents a new method for the separation of merged text/form-structure components in forms. The technique described uses a modified version of the Hough transform to detect the structure of the form. The closed contours of the connected components are approximated by piecewise linear line segments. The parameters of the Hesse normal form of each line segment serve as input for the Hough transform. Compared to the vectorized boundary of characters, the lines of the form structure consist of appreciable more line segments with the same orientation and distance. So, the problem of the form structure detection in the database of line segments can be reduced to the detection of local peaks in the Hough space. Subsequent processing steps reconstruct the remaining contour fragments to characters.<<ETX>>","PeriodicalId":34917,"journal":{"name":"模式识别与人工智能","volume":"1 1","pages":"268-271"},"PeriodicalIF":0.0,"publicationDate":"1992-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73007081","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 : 1992-08-30DOI: 10.1109/ICPR.1992.202149
A. Biancardi, M. Mosconi
This paper proposes a novel approach to program development for highly parallel architectures, primarily as far as debugging is concerned. The visual nature of the debugging stage, when dealing with image-processing algorithms, is heavily supported so that all the relevant information, which is generally either hidden or presented without its logical structures, is made available to programmers. The authors present the modular and portable software system built, in Pavia University, for the PAPIA2 machine.<>
{"title":"Visual debugging for a pyramidal machine","authors":"A. Biancardi, M. Mosconi","doi":"10.1109/ICPR.1992.202149","DOIUrl":"https://doi.org/10.1109/ICPR.1992.202149","url":null,"abstract":"This paper proposes a novel approach to program development for highly parallel architectures, primarily as far as debugging is concerned. The visual nature of the debugging stage, when dealing with image-processing algorithms, is heavily supported so that all the relevant information, which is generally either hidden or presented without its logical structures, is made available to programmers. The authors present the modular and portable software system built, in Pavia University, for the PAPIA2 machine.<<ETX>>","PeriodicalId":34917,"journal":{"name":"模式识别与人工智能","volume":"25 1","pages":"137-141"},"PeriodicalIF":0.0,"publicationDate":"1992-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78245735","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 : 1992-08-30DOI: 10.1109/ICPR.1992.201877
V. Neagoe
Proposes the use of Legendre descriptors (LDs) as features for classification of polygonal closed curves. The normalized cumulative angular function of such a curve is expanded in a Legendre polynomial truncated series whose coefficients are used as shape features called the Legendre descriptors (LDs). By considering several examples of polygonal object classification, the computer simulation shows that the LDs lead to significantly better results (increase of interclass distances), by comparison with the classical Fourier descriptors. It seems that the world of Legendre polynomials is more suitable to approximate a polygonal curve than the world of sinusoidal function.<>
{"title":"Legendre descriptors for classification of polygonal closed curves","authors":"V. Neagoe","doi":"10.1109/ICPR.1992.201877","DOIUrl":"https://doi.org/10.1109/ICPR.1992.201877","url":null,"abstract":"Proposes the use of Legendre descriptors (LDs) as features for classification of polygonal closed curves. The normalized cumulative angular function of such a curve is expanded in a Legendre polynomial truncated series whose coefficients are used as shape features called the Legendre descriptors (LDs). By considering several examples of polygonal object classification, the computer simulation shows that the LDs lead to significantly better results (increase of interclass distances), by comparison with the classical Fourier descriptors. It seems that the world of Legendre polynomials is more suitable to approximate a polygonal curve than the world of sinusoidal function.<<ETX>>","PeriodicalId":34917,"journal":{"name":"模式识别与人工智能","volume":"27 1","pages":"717-720"},"PeriodicalIF":0.0,"publicationDate":"1992-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79438316","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 : 1992-08-30DOI: 10.1109/ICPR.1992.201812
C. Chen
Active sonar classification has been a challenging pattern recognition problem for many years mainly due to the complexity of ocean environment. Improvement of sensors and data acquisition can be very costly and can only provide limited improvement in classification. Neural networks are ideally suited to active sonar classification problems with the potential advantages. In the paper, some active sonar data characteristics are presented, and the performances of several feedforward neural networks are evaluated and compared with the traditional nearest neighbor decision rule. It is concluded that the neural networks studied not only can outperform but also are far more robust than the traditional classifiers.<>
{"title":"Neural networks for active sonar classification","authors":"C. Chen","doi":"10.1109/ICPR.1992.201812","DOIUrl":"https://doi.org/10.1109/ICPR.1992.201812","url":null,"abstract":"Active sonar classification has been a challenging pattern recognition problem for many years mainly due to the complexity of ocean environment. Improvement of sensors and data acquisition can be very costly and can only provide limited improvement in classification. Neural networks are ideally suited to active sonar classification problems with the potential advantages. In the paper, some active sonar data characteristics are presented, and the performances of several feedforward neural networks are evaluated and compared with the traditional nearest neighbor decision rule. It is concluded that the neural networks studied not only can outperform but also are far more robust than the traditional classifiers.<<ETX>>","PeriodicalId":34917,"journal":{"name":"模式识别与人工智能","volume":"537 1","pages":"438-440"},"PeriodicalIF":0.0,"publicationDate":"1992-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79643945","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：