项目名称: 图文混合笔输入文档分析与识别的理论与方法研究
项目编号: No.61273269
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 自动化技术、计算机技术
项目作者: 周祥东
作者单位: 中国科学院软件研究所
项目金额: 81万元
中文摘要: 随着平板电脑和数码笔等笔输入设备的广泛应用,用户能够在更大的界面上输入文本、绘制图形和表格,同时笔迹能够被这些设备捕获并保存成联机文档。但伴随着输入的便利性,无约束的自由书写也给联机文档的自动处理带来了挑战。要对联机手写文档进行分析,首先需要将页面分割成文档结构对象,例如文本行和图形,然后再对不同的对象分别进行识别。本项目基于条件随机场(CRF)模型,系统地对复杂版面联机手写文档分析与识别中多种上下文信息的表示与融合方法展开研究,具体内容包括:(1)基于空间上下文的图文分离方法;(2)复杂版面联机手写文档的文本行提取方法;(3)基于高阶半马尔科夫条件随机场(semi-CRF)的手写字符串识别方法;(4)高阶条件随机场的快速推断算法。该项目在前期扎实的工作基础上开展,针对研究中的难点设计了明确可行的技术路线。本研究内容新颖并具有广泛的应用前景和学术价值,研究成果将促进手写设备的推广与应用。
中文关键词: 手写文档;版面分析;文本行提取;手写文本行识别;关键词定位
英文摘要: With the increasing use of pen-based devices, users can input various heterogeneous structures such as text, drawings and table forms freely on a large writing area, with the trajectories captured and saved as ink documents. Nevertheless, accompanied with the convenience of input, the free and heterogeneous structures of ink documents bring new challenges to automatic processing. For ink document analysis, the ink strokes should be first grouped into structural units such as text lines and drawings, which are then recognized respectively. This project systemically investigates the representation and fusion methods of context information in ink document analysis based on conditional random fields (CRFs). The main contents are as follows: (1) text and non-text separation method based on the spatial context; (2) method of text line grouping from ink documents with complex layout; (3) character string recognition method based on high-order semi-Markov conditional random fields; (4) fast inference methods for high-order conditional random fields. This project will be conducted on the basis of our previous research, and most of the technology difficulties have been comprehensively evaluated. The research results will promote the application of pen-based devices。
英文关键词: handwritten documents;document layout analysis;text line extraction;handwritten text recognition;keyword spotting