测绘通报 ›› 2020, Vol. 0 ›› Issue (8): 44-49.doi: 10.13474/j.cnki.11-2246.2020.0246

• 学术研究 • 上一篇    下一篇

SIFT和SVD相结合的GF-2影像感知哈希算法

王昊1,2,3, 张黎明1,2,3,4, 慕号伟1,2,3, 张鑫港1,2,3, 陈金萍1,2,3   

  1. 1. 兰州交通大学测绘与地理信息学院, 甘肃 兰州 730070;
    2. 地理国情监测技术应用国家地方联合工程研究中心, 甘肃 兰州 730070;
    3. 甘肃省地理国情检测工程实验室, 甘肃 兰州 730070;
    4. 扬州微地图地理信息科技有限公司, 江苏 扬州 225002
  • 收稿日期:2019-11-12 发布日期:2020-09-01
  • 通讯作者: 张黎明。E-mail:zhanglm8@gmail.com E-mail:zhanglm8@gmail.com
  • 作者简介:王昊(1996-),男,硕士生,主要从事高分辨率遥感影像感知哈希研究。E-mail:lzjtu.wh@gmail.com
  • 基金资助:
    国家自然科学资金(41761080);兰州市人才创新创业科技计划(2016-RC-59);甘肃高等学校产业支撑引导项目(2019C-04);兰州交通大学优秀平台支持(201806)

Perceptual Hash algorithm for GF-2 image using SIFT and SVD

WANG Hao1,2,3, ZHANG Liming1,2,3,4, MU Haowei1,2,3, ZHANG Xingang1,2,3, CHEN Jinping1,2,3   

  1. 1. Faculty of Geomatics, Lanzhou Jiaotong University, Lanzhou 730070, China;
    2. National-Local Joint Engineering Research Center of Technologies and Applications for National Geographic State Monitoring, Lanzhou 730070, China;
    3. Gansu Provincial Engineering Laboratory for National Geographic State Monitoring, Lanzhou 730070, China;
    4. Yangzhou Micro Map Geographic Information Technology Co., Ltd., Yangzhou 225002, China
  • Received:2019-11-12 Published:2020-09-01

摘要: 为解决GF-2影像在传输、存储及实际使用中面临的局部篡改问题,本文提出了一种用于GF-2影像完整性认证的感知哈希算法。首先采用固定分块的方式,将整幅影像分割成多个子块,运用SIFT算子提取子块的关键点特征信息;然后采用SVD方法提取SIFT关键点特征信息的特征值,选取其中较大的10%特征值构建子块的感知哈希序列;最后串联所有子块的感知哈希序列生成整幅影像的感知哈希值。在影像认证过程中,通过计算原始影像与待认证影像的感知哈希值的差异,从而实现对影像内容的认证。试验表明,该算法能够实现GF-2影像中地物篡改的检测与定位,同时对LSB水印嵌入、DAT格式转换及BMP格式转换等内容保持操作具有良好的稳健性,为GF-2影像的安全使用提供保障。

关键词: SIFT, SVD, GF-2影像, 感知哈希, 内容认证

Abstract: A perceptual Hash algorithm for the GF-2 image is proposed to solve the problem of the local tampering in the transmission, storage and use of the GF-2 image. Firstly, the entire image is divided into multiple sub-blocks by means of fixed partitioning. SIFT operator is used to extract the feature information of key points in sub-blocks. Then, SIFT key feature information is extracted via SVD and the first 10% of them is chosen to construct the perceptual hash of sub-blocks. Finally, the perceptual Hash sequence of all sub-blocks is concatenated to generate the perceptual hash sequence of the whole image. In the process of image authentication, the difference between the original GF-2 image and the image to be authenticated is used for authentication of the content. In the meantime, the algorithm has realized detection and location of the object tampering in the GF-2 image. Experiments show that the algorithm is robust at the operation of LSB watermark embedding, DAT format conversion and BMP format conversion. In a word, the proposed algorithm can improved security for the use of the GF-2 image.

Key words: SIFT, SVD, GF-2 image, perceptual Hash algorithm, content authentication

中图分类号: