李明楚

个人信息Personal Information

教授

博士生导师

硕士生导师

主要任职:Director of Academic Committee at Kaifa District

其他任职:开发区校区学术分委员会主任(Director of Academic Committee at Kaifa Campus)

性别:男

毕业院校:多伦多大学

学位:博士

所在单位:软件学院、国际信息与软件学院

学科:软件工程. 运筹学与控制论

办公地点:开发区(Kaifa District Campus)

联系方式:mingchul@dlut.edu.cn

电子邮箱:mingchul@dlut.edu.cn

扫描关注

论文成果

当前位置: 中文主页 >> 科学研究 >> 论文成果

Secure variable-capacity self-recovery watermarking scheme

点击次数:

论文类型:期刊论文

发表时间:2017-03-01

发表刊物:MULTIMEDIA TOOLS AND APPLICATIONS

收录刊物:SCIE、EI、Scopus

卷号:76

期号:5

页面范围:6941-6972

ISSN号:1380-7501

关键字:Self-recovery watermark; Security; Tamper detection; Variable-capacit; Compound watermark; Three level secret-key embedding scheme (TLSES)

摘要:Since existing watermarking schemes usually cannot recover the tampered position, a secure variable-capacity self-recovery watermarking scheme is proposed. Both watermark embedding capacity and security are taken into account. The original image is divided into texture blocks and smooth blocks, and the texture blocks not only save traditional information, and save the "details" information. The so-called "details" information refers to the texture information, which not only can effectively resist mean attack, but also help to improve the quality of the recovered image to meet the needs of practical work. And then according to the characteristics of different blocks, the different length compound watermarks are produced. The so-called "compound watermarks" include the authentication watermarks and information watermarks. Authentication watermarks are used to detect the tampered region, and the information watermarks which include basic watermark and additional watermark are used to recover image. Then the compound watermarks are inserted into the other blocks based on the new proposed scheme called three level secret-key embedding scheme (TLSES). And then detect the tamper blocks and recover them by the three level tamper detection scheme (TLTDS). The experimental results show that the paper can not only accurately detect the tamper region and recover image, but also can effectively resist mean attack and collage attack.