基于零树编码的票据图像压缩编码算法  

A Compression Coding Algorithm Based on Zerotrees Coding for Bill Image

作　　者：赵法林[1] 李飞[1] 鲍旭东[1] ZHAO Fa-lin,LI Fei,BAO Xu-dong（Lab of Image Science and Technology,Southeast University,Nanjing 210096,China）

机构地区：[1]东南大学影像科学与技术实验室,江苏南京210096

出　　处：《电脑知识与技术》2013年第6期3856-3858,共3页Computer Knowledge and Technology

摘　　要：票据图像是一种广泛应用于人们生活及工作的图像,由于票据图像中主要包含文字、数字及线条等信息,其细节和边缘信息极其重要。一般的压缩方法,如JPEG压缩,已经不适用于票据图像的压缩,会造成边缘细节信息的损失。因此针对票据图像的特殊性,经过颜色聚类和多分辨率分解后,应用改进的零树编码进行压缩编码。实验结果表明,该文方法比JPEG压缩标准具有更高的压缩比和更好的图像阅读质量。The bill image is an image that is widely used in human life and work.Because there is important information in bill image such as words,numbers,and lines,the detail and edge information is extremely important for bill image.General compres sion methods,such as JPEG compression standard,don’t apply to compress bill image,which can cause loss of detail and edge in formation.According to the particularity of bill image,the proposed method applies to compress bill image.Firstly,the color clus tering method and multiresolution decomposition classify bill image to the pyramid structure.And the improved zerotrees coding is used to compress the pyramid structure.The results demonstrated much higher compression rates and better reading quality of the proposed method than of JPEG compression standard.

关 键 词：票据图像 多分辨率分解 零树编码 JPEG压缩标准 边缘细节信息 

分 类 号：TP18[自动化与计算机技术—控制科学与工程;自动化与计算机技术—控制理论与控制工程]