公开(公告)号：US08363950B2

公开(公告)日：2013-01-29

申请号：US13426427

申请日：2012-03-21

申请人： Xinjian Chen ,  Dongmei Zhang ,  Yu Zou ,  Ming Chang ,  Shi Han ,  Jian Wang

发明人： Xinjian Chen ,  Dongmei Zhang ,  Yu Zou ,  Ming Chang ,  Shi Han ,  Jian Wang

IPC分类号： G06K9/00 ,  G06F17/00

CPC分类号： G06K9/00973 ,  G06K9/6292 ,  G06K9/6296

摘要： Described is a technology by which online recognition of handwritten input data is combined with offline recognition and processing to obtain a combined recognition result. In general, the combination improves overall recognition accuracy. In one aspect, online and offline recognition is separately performed to obtain online and offline character-level recognition scores for candidates (hypotheses). A statistical analysis-based combination algorithm, an AdaBoost algorithm, and/or a neural network-based combination may determine a combination function to combine the scores to produce a result set of one or more results. Online and offline radical-level recognition may be performed. For example, a HMM recognizer may generate online radical scores used to build a radical graph, which is then rescored using the offline radical recognition scores. Paths in the rescored graph are then searched to provide the combined recognition result, e.g., corresponding to the path with the highest score.

公开(公告)号：US20080205761A1

公开(公告)日：2008-08-28

申请号：US11680566

申请日：2007-02-28

申请人： Shi Han ,  Yu Zou ,  Ming Chang ,  Peng Liu ,  Yi-Jian Wu ,  Lei Ma ,  Frank Soong ,  Dongmei Zhang ,  Jian Wang

发明人： Shi Han ,  Yu Zou ,  Ming Chang ,  Peng Liu ,  Yi-Jian Wu ,  Lei Ma ,  Frank Soong ,  Dongmei Zhang ,  Jian Wang

IPC分类号： G06K9/18

CPC分类号： G06K9/6807 ,  G06K9/468 ,  G06K2209/01 ,  G06K2209/011

摘要： Exemplary techniques are described for selecting radical sets for use in probabilistic East Asian character recognition algorithms. An exemplary technique includes applying a decomposition rule to each East Asian character of the set to generate a progressive splitting graph where the progressive splitting graph comprises radicals as nodes, formulating an optimization problem to find an optimal set of radicals to represent the set of East Asian characters using maximum likelihood and minimum description length and solving the optimization problem for the optimal set of radicals. Another exemplary technique includes selecting an optimal set of radicals by using a general function that characterizes a radical with respect to other East Asian characters and a complex function that characterizes complexity of a radical.

摘要翻译： 描述了用于选择在概率东亚字符识别算法中使用的激进集合的示例性技术。 一个示例性的技术包括将分解规则应用于集合的每个东亚字符以生成逐行分割图，其中渐进分割图包括基数作为节点，制定优化问题以找到最佳的一组基团以表示东亚集 字符使用最大似然和最小描述长度，并解决优化问题的最佳组的自由基。 另一个示例性技术包括通过使用表征相对于其他东亚字符的基数的一般函数和表征激进的复杂度的复杂函数来选择最佳的自由基集合。

公开(公告)号：US08463043B2

公开(公告)日：2013-06-11

申请号：US13526236

申请日：2012-06-18

申请人： Yu Zou ,  Ming Chang ,  Shi Han ,  Dongmei Zhang ,  Jian Wang

发明人： Yu Zou ,  Ming Chang ,  Shi Han ,  Dongmei Zhang ,  Jian Wang

IPC分类号： G06K9/00 ,  G06K9/46

CPC分类号： G06K9/00416 ,  G06K2209/011

摘要： An exemplary method for online character recognition of characters includes acquiring time sequential, online ink data for a handwritten character, conditioning the ink data to produce conditioned ink data where the conditioned ink data includes information as to writing sequence of the handwritten character and extracting features from the conditioned ink data where the features include a tangent feature, a curvature feature, a local length feature, a connection point feature and an imaginary stroke feature. Such a method may determine neighborhoods for ink data and extract features for each neighborhood. An exemplary character recognition system may use various exemplary methods for training and character recognition.

摘要翻译： 用于字符的在线字符识别的示例性方法包括获取用于手写字符的时间顺序在线墨水数据，调节墨水数据以产生经调节的墨水数据，其中经调节的墨水数据包括关于写入手写字符的序列的信息并从 调节的油墨数据，其中特征包括切线特征，曲率特征，局部长度特征，连接点特征和假想笔划特征。 这种方法可以确定墨水数据的邻域并提取每个邻域的特征。 示例性字符识别系统可以使用用于训练和字符识别的各种示例性方法。

公开(公告)号：US08160362B2

公开(公告)日：2012-04-17

申请号：US13090242

申请日：2011-04-19

申请人： Xinjian Chen ,  Dongmei Zhang ,  Yu Zou ,  Ming Chang ,  Shi Han ,  Jian Wang

发明人： Xinjian Chen ,  Dongmei Zhang ,  Yu Zou ,  Ming Chang ,  Shi Han ,  Jian Wang

IPC分类号： G06K9/00 ,  G06F17/00

CPC分类号： G06K9/00973 ,  G06K9/6292 ,  G06K9/6296

摘要： Described is a technology by which online recognition of handwritten input data is combined with offline recognition and processing to obtain a combined recognition result. In general, the combination improves overall recognition accuracy. In one aspect, online and offline recognition is separately performed to obtain online and offline character-level recognition scores for candidates (hypotheses). A statistical analysis-based combination algorithm, an AdaBoost algorithm, and/or a neural network-based combination may determine a combination function to combine the scores to produce a result set of one or more results. Online and offline radical-level recognition may be performed. For example, a HMM recognizer may generate online radical scores used to build a radical graph, which is then rescored using the offline radical recognition scores. Paths in the rescored graph are then searched to provide the combined recognition result, e.g., corresponding to the path with the highest score.

摘要翻译： 描述了通过在线识别手写输入数据与离线识别和处理相结合以获得组合识别结果的技术。 通常，该组合提高了整体识别精度。 在一个方面，单独执行在线和离线识别以获得用于候选者（假设）的在线和离线角色级识别分数。 基于统计分析的组合算法，AdaBoost算法和/或基于神经网络的组合可以确定组合函数以组合分数以产生一个或多个结果的结果集。 可以执行在线和离线激进级别识别。 例如，HMM识别器可以生成用于构建激进图形的在线激进分数，然后使用离线激进识别分数进行重新分类。 然后，搜索折叠图中的路径以提供组合识别结果，例如对应于具有最高分数的路径。

公开(公告)号：US20110229038A1

公开(公告)日：2011-09-22

申请号：US13118045

申请日：2011-05-27

申请人： Yu Zou ,  Ming Chang ,  Shi Han ,  Dongmei Zhang ,  Jian Wang

发明人： Yu Zou ,  Ming Chang ,  Shi Han ,  Dongmei Zhang ,  Jian Wang

IPC分类号： G06K9/18

CPC分类号： G06K9/00416 ,  G06K2209/011

摘要： An exemplary method for online character recognition of East Asian characters includes acquiring time sequential, online ink data for a handwritten East Asian character, conditioning the ink data to produce conditioned ink data where the conditioned ink data includes information as to writing sequence of the handwritten East Asian character and extracting features from the conditioned ink data where the features include a tangent feature, a curvature feature, a local length feature, a connection point feature and an imaginary stroke feature. Such a method may determine neighborhoods for ink data and extract features for each neighborhood. An exemplary Hidden Markov Model based character recognition system may use various exemplary methods for training and character recognition.

摘要翻译： 用于东亚字符的在线字符识别的示例性方法包括获取用于手写东亚字符的时间顺序在线墨水数据，调节墨水数据以产生经调节的墨水数据，其中调节的墨水数据包括关于写入东方手写的顺序的信息 亚洲字符和从调节的墨水数据中提取特征，其中特征包括切线特征，曲率特征，局部长度特征，连接点特征和假想笔划特征。 这种方法可以确定墨水数据的邻域并提取每个邻域的特征。 基于示例性的基于隐马尔可夫模型的角色识别系统可以使用用于训练和角色识别的各种示例性方法。

公开(公告)号：US07903877B2

公开(公告)日：2011-03-08

申请号：US11682722

申请日：2007-03-06

申请人： Shi Han ,  Yu Zou ,  Ming Chang ,  Peng Liu ,  Yi-Jian Wu ,  Lei Ma ,  Frank Soong ,  Dongmei Zhang ,  Jian Wang

发明人： Shi Han ,  Yu Zou ,  Ming Chang ,  Peng Liu ,  Yi-Jian Wu ,  Lei Ma ,  Frank Soong ,  Dongmei Zhang ,  Jian Wang

IPC分类号： G06K9/00 ,  G06K9/18

CPC分类号： G06K9/00879

摘要： Exemplary methods, systems, and computer-readable media for developing, training and/or using models for online handwriting recognition of characters are described. An exemplary method for building a trainable radical-based HMM for use in character recognition includes defining radical nodes, where a radical node represents a structural element of an character, and defining connection nodes, where a connection node represents a spatial relationship between two or more radicals. Such a method may include determining a number of paths in the radical-based HMM using subsequence direction histogram vector (SDHV) clustering and determining a number of states in the radical-based HMM using curvature scale space-based (CSS) corner detection.

摘要翻译： 描述用于开发，训练和/或使用用于字符的在线手写识别的模型的示例性方法，系统和计算机可读介质。 用于构建用于字符识别的可训练基于激进的基于HMM的示例性方法包括定义基本节点，其中基本节点表示字符的结构元素，并且定义连接节点，其中连接节点表示两个或更多个之间的空间关系 激进分子 这种方法可以包括使用子序列方向直方图向量（SDHV）聚类确定基于激进的HMM中的路径数量，并使用基于曲率空间的（CSS）角检测确定基于激进的HMM中的状态数。

公开(公告)号：US20120251006A1

公开(公告)日：2012-10-04

申请号：US13526236

申请日：2012-06-18

申请人： Yu Zou ,  Ming Chang ,  Shi Han ,  Dongmei Zhang ,  Jian Wang

发明人： Yu Zou ,  Ming Chang ,  Shi Han ,  Dongmei Zhang ,  Jian Wang

IPC分类号： G06K9/46

CPC分类号： G06K9/00416 ,  G06K2209/011

摘要： An exemplary method for online character recognition of characters includes acquiring time sequential, online ink data for a handwritten character, conditioning the ink data to produce conditioned ink data where the conditioned ink data includes information as to writing sequence of the handwritten character and extracting features from the conditioned ink data where the features include a tangent feature, a curvature feature, a local length feature, a connection point feature and an imaginary stroke feature. Such a method may determine neighborhoods for ink data and extract features for each neighborhood. An exemplary character recognition system may use various exemplary methods for training and character recognition.

摘要翻译： 用于字符的在线字符识别的示例性方法包括获取用于手写字符的时间顺序在线墨水数据，调节墨水数据以产生经调节的墨水数据，其中经调节的墨水数据包括关于写入手写字符的序列的信息并从 调节的油墨数据，其中特征包括切线特征，曲率特征，局部长度特征，连接点特征和假想笔画特征。 这种方法可以确定墨水数据的邻域并提取每个邻域的特征。 示例性字符识别系统可以使用用于训练和字符识别的各种示例性方法。

公开(公告)号：US07953279B2

公开(公告)日：2011-05-31

申请号：US11823644

申请日：2007-06-28

申请人： Xinjian Chen ,  Dongmei Zhang ,  Yu Zou ,  Ming Chang ,  Shi Han ,  Jian Wang

发明人： Xinjian Chen ,  Dongmei Zhang ,  Yu Zou ,  Ming Chang ,  Shi Han ,  Jian Wang

IPC分类号： G06K9/00 ,  G06F17/00

CPC分类号： G06K9/00973 ,  G06K9/6292 ,  G06K9/6296

摘要： Described is a technology by which online recognition of handwritten input data is combined with offline recognition and processing to obtain a combined recognition result. In general, the combination improves overall recognition accuracy. In one aspect, online and offline recognition is separately performed to obtain online and offline character-level recognition scores for candidates (hypotheses). A statistical analysis-based combination algorithm, an AdaBoost algorithm, and/or a neural network-based combination may determine a combination function to combine the scores to produce a result set of one or more results. Online and offline radical-level recognition may be performed. For example, a HMM recognizer may generate online radical scores used to build a radical graph, which is then rescored using the offline radical recognition scores. Paths in the rescored graph are then searched to provide the combined recognition result, e.g., corresponding to the path with the highest score.

摘要翻译： 描述了通过在线识别手写输入数据与离线识别和处理相结合以获得组合识别结果的技术。 通常，该组合提高了整体识别精度。 在一个方面，单独执行在线和离线识别以获得用于候选者（假设）的在线和离线角色级识别分数。 基于统计分析的组合算法，AdaBoost算法和/或基于神经网络的组合可以确定组合函数以组合分数以产生一个或多个结果的结果集。 可以执行在线和离线激进级别识别。 例如，HMM识别器可以生成用于构建激进图形的在线激进分数，然后使用离线激进识别分数进行重新分类。 然后，搜索折叠图中的路径以提供组合识别结果，例如对应于具有最高分数的路径。

公开(公告)号：US20080219556A1

公开(公告)日：2008-09-11

申请号：US11682722

申请日：2007-03-06

申请人： Shi Han ,  Yu Zou ,  Ming Chang ,  Peng Liu ,  Yi-Jian Wu ,  Lei Ma ,  Frank Soong ,  Dongmei Zhang ,  Jian Wang

发明人： Shi Han ,  Yu Zou ,  Ming Chang ,  Peng Liu ,  Yi-Jian Wu ,  Lei Ma ,  Frank Soong ,  Dongmei Zhang ,  Jian Wang

IPC分类号： G06K9/18

CPC分类号： G06K9/00879

摘要： Exemplary methods, systems, and computer-readable media for developing, training and/or using models for online handwriting recognition of characters are described. An exemplary method for building a trainable radical-based HMM for use in character recognition includes defining radical nodes, where a radical node represents a structural element of an character, and defining connection nodes, where a connection node represents a spatial relationship between two or more radicals. Such a method may include determining a number of paths in the radical-based HMM using subsequence direction histogram vector (SDHV) clustering and determining a number of states in the radical-based HMM using curvature scale space-based (CSS) corner detection.

摘要翻译： 描述用于开发，训练和/或使用用于字符的在线手写识别的模型的示例性方法，系统和计算机可读介质。 用于构建用于字符识别的可训练基于激进的基于HMM的示例性方法包括定义基本节点，其中基本节点表示字符的结构元素，并且定义连接节点，其中连接节点表示两个或更多个之间的空间关系 激进分子 这种方法可以包括使用子序列方向直方图向量（SDHV）聚类确定基于激进的HMM中的路径数量，并使用基于曲率空间的（CSS）角检测确定基于激进的HMM中的状态数。

公开(公告)号：US20120183223A1

公开(公告)日：2012-07-19

申请号：US13426427

申请日：2012-03-21

申请人： Xinjian Chen ,  Dongmei Zhang ,  Yu Zou ,  Ming Chang ,  Shi Han ,  Jian Wang

发明人： Xinjian Chen ,  Dongmei Zhang ,  Yu Zou ,  Ming Chang ,  Shi Han ,  Jian Wang

IPC分类号： G06K9/62

CPC分类号： G06K9/00973 ,  G06K9/6292 ,  G06K9/6296

摘要： Described is a technology by which online recognition of handwritten input data is combined with offline recognition and processing to obtain a combined recognition result. In general, the combination improves overall recognition accuracy. In one aspect, online and offline recognition is separately performed to obtain online and offline character-level recognition scores for candidates (hypotheses). A statistical analysis-based combination algorithm, an AdaBoost algorithm, and/or a neural network-based combination may determine a combination function to combine the scores to produce a result set of one or more results. Online and offline radical-level recognition may be performed. For example, a HMM recognizer may generate online radical scores used to build a radical graph, which is then rescored using the offline radical recognition scores. Paths in the rescored graph are then searched to provide the combined recognition result, e.g., corresponding to the path with the highest score.

摘要翻译： 描述了通过在线识别手写输入数据与离线识别和处理相结合以获得组合识别结果的技术。 通常，该组合提高了整体识别精度。 在一个方面，单独执行在线和离线识别以获得用于候选者（假设）的在线和离线角色级识别分数。 基于统计分析的组合算法，AdaBoost算法和/或基于神经网络的组合可以确定组合函数以组合分数以产生一个或多个结果的结果集。 可以执行在线和离线激进级别识别。 例如，HMM识别器可以生成用于构建激进图形的在线激进分数，然后使用离线激进识别分数进行重新分类。 然后，搜索折叠图中的路径以提供组合识别结果，例如对应于具有最高分数的路径。