摘要
随着手机等相关便携高分辨率拍照设备的大范围普及,如何有效的从图像中提取文本信息正受到越来越多的重视。文章系统分析了自然取景后图像文字识别中可能面临的问题,对相关重点环节的可用算法流程进行了梳理,讨论了自然光照条件下去噪及倾斜校正等方法的实现。测试实验表明,设计的轻量化系统在易用性前提下具有较好的识别效果。
With the widespread popularization of related portable and high-resolution camera devices such as mobile phones,how to extract effectively text information from images is receiving more and more attention.This paper analyzes systematically the possible problems in image text recognition after natural shooting,sorts out the available algorithm flow of relevant key links,and discusses the realization methods of noise reduction and tilt correction under the natural light conditions.The test experiments show that the designed lightweight system has a better recognition effect under the premise of ease of use.
作者
冯志昕
杨景皓
李鑫龙
周翔
王震
FENG Zhixin;YANG Jinghao;LI Xinlong;ZHOU Xiang;WANG Zhen(China University of Mining and Technology-Beijing,Beijing 100083,China)
出处
《现代信息科技》
2022年第16期26-28,共3页
Modern Information Technology
基金
中国矿业大学(北京)大学生创新训练项目(202104052)
中央高校基本科研业务费专项资金。
关键词
光学字符识别
灰度化处理
二值化处理
降噪处理
倾斜校正
OCR
grayscale processing
binarized processing
noise reduction processing
tilt correction
