基于结构理解的手绘草图表格识别  

Freehand Sketch Table Recognition Based on Structure Understanding

作　　者：盖鑫 黄进[1] 王丹琳 刘杰[1] 田丰[1] 戴国忠[1] Ge Xin;Huang Jin;Wang Danlin;Liu Jie;Tian Feng;Dai Guozhong(Beijing Key Laboratory of Human-Computer Interaction,Institute of Software,Chinese Academy of Sciences,Beijing 100190;School of Computer Science and Technology,University of Chinese Academy of Sciences,Beijing 100049)

机构地区：[1]中国科学院软件研究所人机交互北京市重点实验室,北京100190 [2]中国科学院大学计算机科学与技术学院,北京100049

出　　处：《计算机辅助设计与图形学学报》2024年第12期2051-2068,共18页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金(62172397);中国科学院青年创新促进会项目(2020113)。

摘　　要：基于笔交互的手绘草图表格比基于WIMP界面范式的传统电子表格更易于促进用户思维交流和创造性交互工作,同时可以避免用户学习成本高和交互界面烦琐等问题.然而,由于缺乏专门用于手绘草图表格识别的开源数据集,以及草图本身具有的模糊性、抽象性和用户在绘制草图时的随意性,手绘草图表格的识别仍然面临着巨大的挑战.为了解决存在复杂单元格、结构框线与内容重叠、笔迹重描补笔等具体挑战,提出了一种基于结构理解的手绘草图表格识别算法,通过一个支持向量机(support vector machines,SVM)对笔划信息进行划分,再使用多峰值检测算法与真实交点找寻和单元格特征属性检测结合,实现复杂草图表格结构的识别.客观评估结果显示,与光学字符识别(optical character recognition,OCR)基线算法相比,基于结构理解的手绘草图表格识别算法在面向草图表格结构识别任务中的树编辑距离相似度(tree-edit-distance-based similarity,TEDS)指标提升了13%以上;专家评估结果显示,基于结构理解的手绘草图表格识别算法在表格结构识别和将内容匹配到对应单元格2个评价维度中的识别效果均优于OCR基线算法的识别效果.ion of the sketches themselves and the arbitrariness of the users in drawing them.In order to solve the specific challenges such as the existence of complex cells,the overlapping of structural frame lines and content,and the redrawing and filling of handwriting,propose an algorithm for freehand sketch table recognition based on table structure understand-ing.A support vector machines(SVM)is used to classify the stroke information,and then the multi-peak detection algorithm is combined with the real intersection finding and cell feature attribute detection con-structed in this paper to achieve the recognition of complex sketch table structure.The objective evaluation results show that the tree-edit-distance-based similarity(TEDS)of this algorithm in the sketch-oriented table structure recognition task is improved by more than 13%compared with the OCR baseline algorithm;the expert evaluation results show that this algorithm outperforms the OCR baseline algorithm in the evaluation dimensions of table structure recognition and matching the content to the corresponding cells.

关 键 词：手绘草图 草图结构理解 表格识别算法 人机交互 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]