Abstract:

Due to the interference of transmission medium, the noise pollutes the remote sensing images, which will affect the further processing and application of images. Since the remote sensing images contain rich information, the ordinary computers need to work together to complete the noise reduction task. What’s more, the cloud platform is usually untrustworthy or even malicious. Thus, the traditional image-denoising scheme cannot be directly applied to the cloud platform. In order to ensure the availability and security of remote sensing images, a secure outsourced denoising scheme in remote sensing images on multi-cloud architecture is proposed, which can not only achieve disaster recovery, but also achieve distributed noise reduction and reduce the computational burden of each cloud platform. It uses a [(k,n)] threshold secret sharing scheme to share remote sensing images among multiple cloud servers and these cloud servers denoise the images in the ciphertext domain according to the improved non-local mean noise reduction algorithm.

Key words: remote sensing image, secret sharing, non-local means, secure outsource

摘要：

由于传输介质的干扰，遥感图像会被噪声污染，影响图像的处理和应用。遥感图像包含丰富的信息，致使普通计算机无法独立完成降噪任务，而云平台往往是不可信任甚至是恶意的，因此传统的图像降噪方案无法直接应用到云平台上。为保障遥感图像的可用性与安全性，提出一种基于多云协同的遥感图像安全外包降噪方案，既能实现灾备，又能实现分布式降噪，减少每个云平台的计算负担。利用[(k,n)]门限秘密共享方案将遥感图像共享给多个云服务器，云服务器根据改进的非局部均值降噪算法在密文域中完成图像降噪。

关键词: 遥感图像, 秘密共享, 非局部均值, 安全外包