基于稀疏重建和总变分的模糊文本图像复原方法  

Restoration Method of Fuzzy Text Image Based on Sparse Reconstruction and Total Variation

作　　者：彭姣丽 张人杰 郭李平[3] PENG Jiaoli;ZHANG Renjie;GUO Liping(Hunan Polytechnic of Environment and Biology,Hengyang,Hunan,China 421005;Hunan Post and Telecommunication College,Changsha,Hunan,China 410015;Hunan Industry Polytechnic,Changsha,Hunan,China 410017)

机构地区：[1]湖南环境生物职业技术学院,湖南衡阳421005 [2]湖南邮电职业技术学院,湖南长沙410015 [3]湖南工业职业技术学院,湖南长沙410017

出　　处：《湖南邮电职业技术学院学报》2023年第4期1-5,共5页Journal of Hunan Post and Telecommunication College

基　　金：2023年衡阳市社会科学基金项目“人工智能时代衡阳地区职业院校计算机类专业人才培养体系研究”(项目编号:2023C035)。

摘　　要：由于图片具有鲜明的色彩、较强的方向性和文字边缘包含重要信息的特点,为了更好地复原模糊的文本图像,在目前已有的图像复原算法的基础上,把文本图像作为整体,针对由运动和噪声引起的文本图像模糊,研究了基于稀疏重建和总变分的文本图像复原方法。实验证明,此方法复原的文本图像效果较好,信噪比高,具有较高的实用价值。Due to the distinct colors,strong directionality,and important information contained in text edges of images,in order to better restore fuzzy text images,based on existing image restoration algorithms,the text image is treated as a whole,and a text image restoration method based on sparse reconstruction and total variation is studied for text image fuzzy caused by motion and noise.And through experiments,it has been proven that the text image restored by this method has good performance,high signal-to-noise ratio,and high practical value.

关 键 词：文本图像 模糊 稀疏 自适应总变分 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]