手机拍照阅读系统的设计与实现  

Design and implementation of mobile photo-reading system

作　　者：费梦圆 闯跃龙[1] 郭文平[1] 

机构地区：[1]台州学院数学与信息工程学院,浙江台州317000

出　　处：《信息技术》2017年第2期29-32,42,共5页Information Technology

基　　金：国家级大学生创新创业训练计划项目(201410350010)

摘　　要：针对视障用户获取文本信息的障碍,设计并实现了一种利用移动设备实现从拍照到文本识别的系统。该系统利用数字图像处理中的基本算法实现图像中的文本区域识别,利用OCR(Optical Character Recognition)引擎实现从图像到文本的转换,最终利用TTS(Text to Speech)和文本中相结合的方式反馈用户。为了提高本中识别的准确性,提出图象旋转矫正算法修正拍照时产生的旋转问题。为了提高系统的执行效率和用户的体验,采用Android系统的多线程技术,对文本中识别与TTS播报进行并行处理。通过实验表明系统完全可以胜任日常文本检测与识别功能,具有广阔的市场前景。To solve the problem ol text acquirement lor vision impaired people,a mobile photo-reading system is designed and implemented. The system makes use ol several algorithms ol digital image processing to detect text regions within images,and introduces OCR （Optical Character Recognition） to recognize text information. Finally,the combination ol text information and TTS is send to users. To raise the accuracy ol text detection,an image derotation algorithm is proposed which is used to correct rotation angle. To improve the efficiency ol the system and user experience, multi-threading technique based on Android platlorm is adopted, which results in the parallel processing ol text recognition and TTS broadcast. The experiment shows that the system is excellent enough to detect and recognize text inlormation,which has the broad market prospect.

关 键 词：ANDROID OCR 文本检测 文本识别 TTS 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]