测绘通报 ›› 2021, Vol. 0 ›› Issue (12): 22-27.doi: 10.13474/j.cnki.11-2246.2021.366

• 学术研究 • 上一篇    下一篇

运用ULBP的高分辨率遥感影像感知哈希完整性认证算法

刘明轩1,2,3, 张黎明1,2,3, 王昊1,2,3, 马文骏1,2,3, 李玉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. 甘肃省地理国情监测工程试验室, 甘肃 兰州 730070
  • 收稿日期:2020-12-07 发布日期:2021-12-30
  • 通讯作者: 张黎明。E-mail:zhanglm8@gmail.com
  • 作者简介:刘明轩(1997-),男,硕士生,主要从事遥感影像感知哈希完整性认证研究。E-mail:13840348310@163.com
  • 基金资助:
    国家自然科学基金(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(201806)

Perceptual Hash integrity authentication algorithm for high-resolution remote sensing image using ULBP

LIU Mingxuan1,2,3, ZHANG Liming1,2,3, WANG Hao1,2,3, MA Wenjun1,2,3, LI Yu1,2,3   

  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
  • Received:2020-12-07 Published:2021-12-30

摘要: 针对现有高分辨率遥感影像感知哈希认证算法中特征提取精度和算法稳健性不能兼顾的问题,本文提出一种运用ULBP的高分辨率遥感影像感知哈希完整性的认证算法。首先对遥感影像进行格网划分,将影像划分为多个子块,运用ULBP算法提取子块的纹理特征;然后计算每个子块内纹理特征的直方图分布,将其结果与均值二值化后得到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。在影像认证时,首先计算原始影像与待检测影像的哈希序列,然后计算两者的感知哈希序列平均汉明距离,完成对高分辨率遥感影像内容的完整性认证与篡改定位。试验表明,该算法能够识别高分辨率遥感影像中地物较为平滑的区域,同时对JPEG压缩、高斯噪声添加、BMP格式转换等操作具有良好的稳健性,为高分辨遥感影像完整性认证提供支持。

关键词: 感知哈希, ULBP, 高分影像, 内容认证, 纹理特征

Abstract: Aiming at the problem that feature extraction accuracy and algorithm robustness cannot be taken into account in the existing high-resolution remote sensing image perception Hash authentication algorithms, this paper proposes a high-resolution remote sensing image perception Hash integrity authentication algorithm using ULBP. Firstly, the remote sensing image was divided into grids, and the image is divided into multiple sub-blocks. ULBP algorithm is used to extract the texture features of sub-blocks. Then, the histogram distribution of texture feature in each sub-block is calculated and the perceived hash sequence of sub-block is obtained after binarization of the result and the mean value. Finally, the perceptive Hash sequence of all sub-blocks is concatenated to generate the perceptive hash value of the whole image. In image authentication, the Hash sequence of the original image and the image to be detected is first calculated, and then the average hamming distance of the perceived Hash sequence of the two is calculated, so as to complete the integrity authentication and tamper location of the high-resolution image content. Experiments show that the algorithm can identify the relatively smooth areas of ground objects in high-resolution remote sensing images, and has good robustness for the maintenance of contents such as JPEG compression, Gaussian noise and BMP format conversion, thus providing support for the integrity authentication of high-resolution remote sensing images.

Key words: perceptual Hash, ULBP, high-resolution image, content authentication, textural features

中图分类号: