Pub Date : 2002-08-06DOI: 10.1109/IWFHR.2002.1030917
G. Leedham, Saket Varma, A. Patankar, V. Govindaraju
Before any processing of the textual content of a document image can be performed the text must be separated from the background of the image. Several thresholding algorithms have previously been proposed and are widely used in document processing. None have been shown effective at thresholding difficult documents where the background and foreground are non-uniform. In this paper we investigate the use of three global thresholding algorithms (Otsu's, Kapur's entropy and Solihin's quadratic integral ratio (QIR)) as the first stage in a multi-stage thresholding algorithm for use in degraded document images. It is concluded that Otsu's and Kapur's algorithms do not work well for difficult documents as they tend to over-threshold the image, thus losing much of the useful information. The QIR algorithm is more accurate in separating the foreground and background in these images, leaving a range of undecided, fuzzy, pixels for later processing in a subsequent stage.
{"title":"Separating text and background in degraded document images - a comparison of global thresholding techniques for multi-stage thresholding","authors":"G. Leedham, Saket Varma, A. Patankar, V. Govindaraju","doi":"10.1109/IWFHR.2002.1030917","DOIUrl":"https://doi.org/10.1109/IWFHR.2002.1030917","url":null,"abstract":"Before any processing of the textual content of a document image can be performed the text must be separated from the background of the image. Several thresholding algorithms have previously been proposed and are widely used in document processing. None have been shown effective at thresholding difficult documents where the background and foreground are non-uniform. In this paper we investigate the use of three global thresholding algorithms (Otsu's, Kapur's entropy and Solihin's quadratic integral ratio (QIR)) as the first stage in a multi-stage thresholding algorithm for use in degraded document images. It is concluded that Otsu's and Kapur's algorithms do not work well for difficult documents as they tend to over-threshold the image, thus losing much of the useful information. The QIR algorithm is more accurate in separating the foreground and background in these images, leaving a range of undecided, fuzzy, pixels for later processing in a subsequent stage.","PeriodicalId":114017,"journal":{"name":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","volume":"129 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116171695","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 : 2002-08-06DOI: 10.1109/IWFHR.2002.1030902
Jue Wang, Chenyu Wu, Ying-Qing Xu, H. Shum, Liang Ji
In this paper an integrated approach for modeling, learning and synthesizing personal cursive handwriting is proposed. Cursive handwriting is modeled by a tri-unit handwriting model, which focuses on both the handwritten letters and the interconnection strokes of adjacent letters. Handwriting strokes are formed from generative models that are based on control points and B-spline curves. In the two-step learning process, a template-based matching algorithm and a data congealing algorithm are first proposed to extract training vectors from handwriting samples, and then letter style models and concatenation style models are trained separately. In the synthesis process, isolated letters and ligature strokes are generated from the learned models and concatenated with each other to produce the whole word trajectory, with guidance from a deformable model. Experimental results show that the proposed system can effectively learn the individual style of cursive handwriting and has the ability to generate novel handwriting of the same style.
{"title":"Learning-based cursive handwriting synthesis","authors":"Jue Wang, Chenyu Wu, Ying-Qing Xu, H. Shum, Liang Ji","doi":"10.1109/IWFHR.2002.1030902","DOIUrl":"https://doi.org/10.1109/IWFHR.2002.1030902","url":null,"abstract":"In this paper an integrated approach for modeling, learning and synthesizing personal cursive handwriting is proposed. Cursive handwriting is modeled by a tri-unit handwriting model, which focuses on both the handwritten letters and the interconnection strokes of adjacent letters. Handwriting strokes are formed from generative models that are based on control points and B-spline curves. In the two-step learning process, a template-based matching algorithm and a data congealing algorithm are first proposed to extract training vectors from handwriting samples, and then letter style models and concatenation style models are trained separately. In the synthesis process, isolated letters and ligature strokes are generated from the learned models and concatenated with each other to produce the whole word trajectory, with guidance from a deformable model. Experimental results show that the proposed system can effectively learn the individual style of cursive handwriting and has the ability to generate novel handwriting of the same style.","PeriodicalId":114017,"journal":{"name":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125379263","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 : 2002-08-06DOI: 10.1109/IWFHR.2002.1030883
Claus Bahlmann, B. Haasdonk, H. Burkhardt
In this paper we describe a novel classification approach for online handwriting recognition. The technique combines dynamic time warping (DTW) and support vector machines (SVMs) by establishing a new SVM kernel. We call this kernel Gaussian DTW (GDTW) kernel. This kernel approach has a main advantage over common HMM techniques. It does not assume a model for the generative class conditional densities. Instead, it directly addresses the problem of discrimination by creating class boundaries and thus is less sensitive to modeling assumptions. By incorporating DTW in the kernel function, general classification problems with variable-sized sequential data can be handled. In this respect the proposed method can be straightforwardly applied to all classification problems, where DTW gives a reasonable distance measure, e.g., speech recognition or genome processing. We show experiments with this kernel approach on the UNIPEN handwriting data, achieving results comparable to an HMM-based technique.
{"title":"Online handwriting recognition with support vector machines - a kernel approach","authors":"Claus Bahlmann, B. Haasdonk, H. Burkhardt","doi":"10.1109/IWFHR.2002.1030883","DOIUrl":"https://doi.org/10.1109/IWFHR.2002.1030883","url":null,"abstract":"In this paper we describe a novel classification approach for online handwriting recognition. The technique combines dynamic time warping (DTW) and support vector machines (SVMs) by establishing a new SVM kernel. We call this kernel Gaussian DTW (GDTW) kernel. This kernel approach has a main advantage over common HMM techniques. It does not assume a model for the generative class conditional densities. Instead, it directly addresses the problem of discrimination by creating class boundaries and thus is less sensitive to modeling assumptions. By incorporating DTW in the kernel function, general classification problems with variable-sized sequential data can be handled. In this respect the proposed method can be straightforwardly applied to all classification problems, where DTW gives a reasonable distance measure, e.g., speech recognition or genome processing. We show experiments with this kernel approach on the UNIPEN handwriting data, achieving results comparable to an HMM-based technique.","PeriodicalId":114017,"journal":{"name":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122022696","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 : 2002-08-06DOI: 10.1109/IWFHR.2002.1030913
L. Vuurpijl, Lambert Schomaker, E. Broek
In this paper, the image retrieval system Vind(x) is described. The architecture of the system and first user-experiences are reported. Using Vind(x), users on the Internet may cooperatively annotate objects in paintings by use of the pen or mouse. The collected data can be searched through query-by-drawing techniques, but can also serve as an (ever-growing) training and benchmark set for the development of automated image retrieval systems of the future. Several other examples of cooperative annotation are presented in order to underline the importance of this concept for the design of pattern recognition systems and the labeling of large quantities of scanned documents or online data.
{"title":"Vind(x): using the user through cooperative annotation","authors":"L. Vuurpijl, Lambert Schomaker, E. Broek","doi":"10.1109/IWFHR.2002.1030913","DOIUrl":"https://doi.org/10.1109/IWFHR.2002.1030913","url":null,"abstract":"In this paper, the image retrieval system Vind(x) is described. The architecture of the system and first user-experiences are reported. Using Vind(x), users on the Internet may cooperatively annotate objects in paintings by use of the pen or mouse. The collected data can be searched through query-by-drawing techniques, but can also serve as an (ever-growing) training and benchmark set for the development of automated image retrieval systems of the future. Several other examples of cooperative annotation are presented in order to underline the importance of this concept for the design of pattern recognition systems and the labeling of large quantities of scanned documents or online data.","PeriodicalId":114017,"journal":{"name":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","volume":"159 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114433885","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 : 2002-08-06DOI: 10.1109/IWFHR.2002.1030957
S. Al-Maadeed, D. Elliman, C. Higgins
In this paper we present a new database for off-line Arabic handwriting recognition, together with associated preprocessing procedures. We have developed a new database for the collection, storage and retrieval of Arabic handwritten text (AHDB). This is an advance both in terms of the size of the database as well as the number of different writers involved. We further designed an innovative, simple yet powerful, in place tagging procedure for our database. It enables us to easily extract the bitmaps of words. We also constructed a preprocessing class, which contains some useful preprocessing operations. In this paper the most popular words in Arabic writing were identified for the first time, using an associated program.
{"title":"A data base for Arabic handwritten text recognition research","authors":"S. Al-Maadeed, D. Elliman, C. Higgins","doi":"10.1109/IWFHR.2002.1030957","DOIUrl":"https://doi.org/10.1109/IWFHR.2002.1030957","url":null,"abstract":"In this paper we present a new database for off-line Arabic handwriting recognition, together with associated preprocessing procedures. We have developed a new database for the collection, storage and retrieval of Arabic handwritten text (AHDB). This is an advance both in terms of the size of the database as well as the number of different writers involved. We further designed an innovative, simple yet powerful, in place tagging procedure for our database. It enables us to easily extract the bitmaps of words. We also constructed a preprocessing class, which contains some useful preprocessing operations. In this paper the most popular words in Arabic writing were identified for the first time, using an associated program.","PeriodicalId":114017,"journal":{"name":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131268630","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 : 2002-08-06DOI: 10.1109/IWFHR.2002.1030955
C. Stefano, A. Marcelli
This paper presents a method for locating the points where most likely joints between successive characters within a word occur. The proposed method, whose basic assumptions follow from handwriting generation studies, relies upon a set of morphological criteria applied to both the ligatures and the terminal regions of successive characters in order to decide the most appropriate position for the segmentation points. It does not exploit any temporal information, but rather it manipulates shape information, thus is suitable for both online and off-line handwriting processing. An experimental procedure, adopted to quantitatively evaluate the performance of the proposed algorithm without using any classification method, is also introduced.
{"title":"From ligatures to characters: a shape-based algorithm for handwriting segmentation","authors":"C. Stefano, A. Marcelli","doi":"10.1109/IWFHR.2002.1030955","DOIUrl":"https://doi.org/10.1109/IWFHR.2002.1030955","url":null,"abstract":"This paper presents a method for locating the points where most likely joints between successive characters within a word occur. The proposed method, whose basic assumptions follow from handwriting generation studies, relies upon a set of morphological criteria applied to both the ligatures and the terminal regions of successive characters in order to decide the most appropriate position for the segmentation points. It does not exploit any temporal information, but rather it manipulates shape information, thus is suitable for both online and off-line handwriting processing. An experimental procedure, adopted to quantitatively evaluate the performance of the proposed algorithm without using any classification method, is also introduced.","PeriodicalId":114017,"journal":{"name":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125683637","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 : 2002-08-06DOI: 10.1109/IWFHR.2002.1030912
G. Seni
This paper presents a novel handwriting input user interface (UI)for small portable devices with a touch-enabled screen. This UI includes a writing area on the screen that behaves as a "treadmill" such that electronic ink input is immediately moved from right to left while it is being entered, giving the user the feeling of writing text on a virtual "ticker-tape". This method allows the user to write continuously without running out of writing space, and takes up very little screen real-estate to implement. The UI communicates with a recognition engine capable of recognizing continuously input handwritten text and that can buffer incomplete ink entries. The UI technique described, unlike prior interfaces in which space constraints limit the ability to continuously write on the device screen and thus slow text input, allows the full throughput benefit of continuous text input to be realized within a very small writing space.
{"title":"Treadmill ink - enabling continuous pen input on small devices","authors":"G. Seni","doi":"10.1109/IWFHR.2002.1030912","DOIUrl":"https://doi.org/10.1109/IWFHR.2002.1030912","url":null,"abstract":"This paper presents a novel handwriting input user interface (UI)for small portable devices with a touch-enabled screen. This UI includes a writing area on the screen that behaves as a \"treadmill\" such that electronic ink input is immediately moved from right to left while it is being entered, giving the user the feeling of writing text on a virtual \"ticker-tape\". This method allows the user to write continuously without running out of writing space, and takes up very little screen real-estate to implement. The UI communicates with a recognition engine capable of recognizing continuously input handwritten text and that can buffer incomplete ink entries. The UI technique described, unlike prior interfaces in which space constraints limit the ability to continuously write on the device screen and thus slow text input, allows the full throughput benefit of continuous text input to be realized within a very small writing space.","PeriodicalId":114017,"journal":{"name":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127752842","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 : 2002-08-06DOI: 10.1109/IWFHR.2002.1030954
D. You, Gyeonghwan Kim
A slant correction method for handwritten Korean strings based on analysis of stroke distribution, which reflects structural properties of Korean characters, is presented in this paper. The method aims to deal with typical problems which have been frequently observed in slant correction of handwritten Korean strings with conventional approaches developed for English/European languages. Extracted strokes from a line of text image are classified into two clusters: vertical and diagonal. Gaussian modeling is applied to each of the clusters and the slant angle is estimated from the model which represents the vertical strokes. Experimental results support the effectiveness of the proposed method. For the performance comparison 1,600 handwritten address sting images were used, and success rate of 96.7%, which is much higher than other conventional approaches, has been achieved.
{"title":"Slant correction of handwritten strings based on structural properties of Korean characters","authors":"D. You, Gyeonghwan Kim","doi":"10.1109/IWFHR.2002.1030954","DOIUrl":"https://doi.org/10.1109/IWFHR.2002.1030954","url":null,"abstract":"A slant correction method for handwritten Korean strings based on analysis of stroke distribution, which reflects structural properties of Korean characters, is presented in this paper. The method aims to deal with typical problems which have been frequently observed in slant correction of handwritten Korean strings with conventional approaches developed for English/European languages. Extracted strokes from a line of text image are classified into two clusters: vertical and diagonal. Gaussian modeling is applied to each of the clusters and the slant angle is estimated from the model which represents the vertical strokes. Experimental results support the effectiveness of the proposed method. For the performance comparison 1,600 handwritten address sting images were used, and success rate of 96.7%, which is much higher than other conventional approaches, has been achieved.","PeriodicalId":114017,"journal":{"name":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127781914","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 : 2002-08-06DOI: 10.1109/IWFHR.2002.1030943
R. Milewski, V. Govindaraju
Artificial Intelligence (AI) plays the following two crucial roles in medical form analysis: recognition, as an input, of the New York State (NYS) Prehospital Care Report (PCR), and data inferences as an output. The PCR provides medical, legal, and quality assurance (QA) data (approximately 2-3 Years behind in storage and analysis) that needs to be efficiently centralized to aid health care. Automating NYS PCR analysis will facilitate a more efficient and useful description of a patient being admitted to a hospital emergency room (ER). ER environments can be highly stressful on the human body given the time constraints of bioterrorism, trauma and/or disease. The recognition task will allow these ER health care professionals to evaluate all data and emergency techniques performed by paramedics and emergency medical technicians (EMT's). A computer screen, presenting diagrams, descriptions and inferences of a human body, representing the patient, will be updated with the corresponding handwritten PCR information. This information can then be transported to a central data bank where other hospitals can determine if there are possible outbreaks due to bio-terrorism, disease, hazardous materials incident or other non-obvious mass casualty incidents (MCI). Currently, it may take several days or even weeks, when it is clearly too late, to discover a massive atrocity. The recognition process will involve a method for reducing the size of the lexicon by integrating semantic knowledge with pattern recognition data.
{"title":"Medical word recognition using a computational semantic lexicon","authors":"R. Milewski, V. Govindaraju","doi":"10.1109/IWFHR.2002.1030943","DOIUrl":"https://doi.org/10.1109/IWFHR.2002.1030943","url":null,"abstract":"Artificial Intelligence (AI) plays the following two crucial roles in medical form analysis: recognition, as an input, of the New York State (NYS) Prehospital Care Report (PCR), and data inferences as an output. The PCR provides medical, legal, and quality assurance (QA) data (approximately 2-3 Years behind in storage and analysis) that needs to be efficiently centralized to aid health care. Automating NYS PCR analysis will facilitate a more efficient and useful description of a patient being admitted to a hospital emergency room (ER). ER environments can be highly stressful on the human body given the time constraints of bioterrorism, trauma and/or disease. The recognition task will allow these ER health care professionals to evaluate all data and emergency techniques performed by paramedics and emergency medical technicians (EMT's). A computer screen, presenting diagrams, descriptions and inferences of a human body, representing the patient, will be updated with the corresponding handwritten PCR information. This information can then be transported to a central data bank where other hospitals can determine if there are possible outbreaks due to bio-terrorism, disease, hazardous materials incident or other non-obvious mass casualty incidents (MCI). Currently, it may take several days or even weeks, when it is clearly too late, to discover a massive atrocity. The recognition process will involve a method for reducing the size of the lexicon by integrating semantic knowledge with pattern recognition data.","PeriodicalId":114017,"journal":{"name":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132324049","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 : 2002-08-06DOI: 10.1109/IWFHR.2002.1030887
S. Uchida, H. Sakoe
A fast elastic matching based handwritten character recognition method is investigated. In the method, an unconstrained elastic matching technique, where the matching is optimized locally and individually on each pixel, is utilized together with its a posteriori evaluation based on the eigen-deformations of handwritten characters. Our experimental results show that high recognition rates can be attained by the present method with feasible computations.
{"title":"A handwritten character recognition method based on unconstrained elastic matching and eigen-deformations","authors":"S. Uchida, H. Sakoe","doi":"10.1109/IWFHR.2002.1030887","DOIUrl":"https://doi.org/10.1109/IWFHR.2002.1030887","url":null,"abstract":"A fast elastic matching based handwritten character recognition method is investigated. In the method, an unconstrained elastic matching technique, where the matching is optimized locally and individually on each pixel, is utilized together with its a posteriori evaluation based on the eigen-deformations of handwritten characters. Our experimental results show that high recognition rates can be attained by the present method with feasible computations.","PeriodicalId":114017,"journal":{"name":"Proceedings Eighth International Workshop on Frontiers in Handwriting Recognition","volume":"80 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131824305","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：