TFR算法的身份证照图像处理方法  

Image processing method for ID photos based on TFR algorithm

作　　者：兰红[1] 黄敏[1] LAN Hong;HUANG Min(School of Information Engineering,Jiangxi University of Science and Technology,Ganzhou 341000,China)

机构地区：[1]江西理工大学信息工程学院,江西赣州341000

出　　处：《江西理工大学学报》2020年第1期77-82,共6页Journal of Jiangxi University of Science and Technology

基　　金：国家自然科学基金资助项目(61762046);江西省教育厅科技重点项目(GJJ160599);江西省自然科学基金(20161BAB212048)。

摘　　要：二代居民身份证对数字相片具有严格标准,为了把不符合标准的相片处理成合格的身份证照,提出了TFR算法的身份证照图像处理方法。TFR算法对不符合标准的相片分三步进行处理:①先用头顶距裁剪T算法裁剪相片;②再用改进FloodFill的F算法进行相片背景替换;③最后再用均值RGB换肤R算法进行肤色调整。通过以上三步即能成功将不合格相片转换成合格的身份证照,并通过"二代居民身份证相片检测平台"检测合格。针对处理证件照文献不足的情况,创新地提出了TFR算法来解决现有问题。运用本文证件照图像处理方法处理相片,用户只需提供一张背景较简单的单人照,就能将之自动转换成符合标准的身份证相片,解决了通常需要前往照相馆拍摄身份证相片所带来的不便。Considering that the standards for the second generation ID photos are strict, a TFR algorithm for image processing of ID photos is proposed in order to process non-conforming photos into qualified ID photos.The TFR algorithm processes the non-conforming photos in three steps. Firstly, cut the photo to fit the scales of requirements with T algorithm of the clipping algorithm based on distance along the head. Secondly, replace the photo background with F algorithm of revised FloodFill algorithm. Thirdly, adjust and equalize the face color of the photo with R algorithm of the mean value RGB skin algorithm. Through the above three steps, the unqualified photos can be successfully converted into qualified ID photos, and can pass the "The second generation ID photo detection platform ". In view of the insufficient references for ID photos, the TFR algorithm is innovatively proposed to solve the existing problems. A photo with a simple background can be processed into a standard ID photo by TFR algorithm, which can avoid the inconvenience of going to the photo studio in person.

关 键 词：证件照处理 头顶距裁剪算法 改进的FloodFill算法 均值RGB换肤算法 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]