基于压缩感知技术的秘密图像分存方案

发布时间：2018-02-03 20:06

本文关键词： 秘密共享图像分存压缩感知稀疏变换信号重构　出处：《内蒙古大学》2017年硕士论文　论文类型：学位论文

【摘要】：传统秘密图像分存技术需要对图像的所有数据进行处理,由于图像数据量大,所以算法执行时间比较长,而且分存算法产生的数据总量与原始图像数据量相比扩张明显,会对网络传输和存储造成较大负担。当图像信号具有稀疏性,或者在某一种稀疏变换基的表示下绝大部分系数为零或近似为零时,压缩感知技术通过构造一个与稀疏基不相关的适合的测量矩阵对原始图像进行感知测量。测量得到的数据包含了原始图像的绝大部分有用信息,在确保能够精确重构图像的同时将原图像从高维压缩到低维,极大地减少了需要处理的数据量,能很好地解决传统方法中由于数据量大而导致的诸多问题。在信号重构端,通过一定的重构算法可以获得原始图像信息一个精确的或者高度近似的逼近。本文将传统秘密图像分存与压缩感知技术相结合,实现了一个基于压缩感知技术的秘密图像分存方案。为了进一步提高方案的性能,又从信号的稀疏表示、信号的感知测量和信号重构三个方面对方案进行了优化。通过实验我们发现本文方案能明显降低需要处理的数据量,有效减少算法执行时间,并且重构图像可以达到一个较理想的视觉效果。实验结果表明:与经典的Thien-Lin方案相比,本文实现的初始方案能平均减少48.66%的图像分存时间和29.32%的图像还原时间。优化方案可以平均减少58.77%的分存时间和58.16%的还原时间,而且优化方案的重构图像精度比初始方案提高了 4-10dB。
[Abstract]:The traditional secret image sharing technology needs to process all the data of the image. Because of the large amount of image data, the execution time of the algorithm is relatively long. Moreover, the total amount of data generated by the split storage algorithm is obviously expanded compared with the original image data volume, which will create a great burden on the network transmission and storage, when the image signal is sparse. Or most of the coefficients are zero or approximately 00:00 under the representation of a sparse transform basis. The compression sensing technique constructs a suitable measurement matrix which is not related to the sparse base to measure the original image. The measured data contain most of the useful information of the original image. At the same time, the original image can be compressed from high dimension to low dimension, which greatly reduces the amount of data to be processed. It can solve many problems caused by large amount of data in traditional methods. A precise or highly approximate approximation of the original image information can be obtained by a certain reconstruction algorithm. In this paper, the traditional secret image sharing and compression sensing techniques are combined. In order to further improve the performance of the scheme, a secret image sharing scheme based on compressed sensing technology is implemented. Through experiments, we find that the proposed scheme can significantly reduce the amount of data to be processed, and effectively reduce the execution time of the algorithm. And the reconstructed image can achieve an ideal visual effect. Experimental results show that: compared with the classical Thien-Lin scheme. The initial scheme in this paper can reduce the average time of image sharing by 48.66% and the time of image restoration by 29.32%. The optimized scheme can reduce the time of sharing by 58.77% and 58.16% on average. The restore time. Moreover, the reconstructed image precision of the optimized scheme is 4-10 dB higher than that of the original scheme.
【学位授予单位】：内蒙古大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【参考文献】