基于机器视觉的英语拼写语法错误识别  

Recognition of English spelling and grammar errors based on machine vision

作　　者：晏芳[1] 胡社利[1] 司海峰[1] YAN Fang;HU Sheli;SI Haifeng(Xi’an Siyuan University,Xi’an 710038,China)

机构地区：[1]西安思源学院,西安710038

出　　处：《自动化与仪器仪表》2022年第12期138-142,共5页Automation & Instrumentation

基　　金：校级科研项目《基于就业岗位群的英语应用能力要素分析与体系构建》(XASY-A1440)。

摘　　要：针对当前机器英语语法错误识别纠正精确度低且模式单一的问题,以Transformer模型为基础提出一款基于机器视觉的英语拼写语法错误识别纠正模型,以提高英语语法纠正的整体水平。实验结果表明,本研究设计的视觉系统能够准确地进行图像英语信息的提取;以Transformer模型作为英语拼写语法错误纠正的基础模型,能够完成英语语法错误的识别纠正,与传统模型相比,Transformer模型的召回率达到了33.21,在F0.5指标上达到了64.56,更加适用于英语语法错误的纠正。以上实验结果验证了本研究选用的Transformer模型的合理性,同时验证了提出的基于机器视觉的英语拼写语法错误识别纠正模型的可行性,具有一定的实际参考价值。In view of the low accuracy of the current machine English grammar error recognition and correction, a machine vision based English spelling grammar error recognition and correction model based on transformer model is proposed to improve the overall level of English grammar correction. The experimental results show that the vision system designed in this study can accurately extract English information from images;Taking transformer model as the basic model for correcting English spelling and grammar errors, it can complete the recognition and correction of English grammar errors. Compared with the traditional model, the recall rate of transformer model reaches 33.21 and the F0.5index reaches 64.56, which is more suitable for correcting English grammar errors. The above experimental results verify the rationality of the transformer model selected in this study, and verify the feasibility of the proposed machine vision based English spelling and grammar error recognition and correction model, which has certain practical reference value.

关 键 词：语法错误纠正 机器视觉 图像处理 Transformer模型 

分 类 号：TP392[自动化与计算机技术—计算机应用技术]