基于参数化流形学习的压缩传感重构方法

压缩传感是一种新的信息获取理论,它突破了传统的采样理论,将数据采集和压缩合二为一,再利用重构算法将原始数据恢复。为了能够得到更好的压缩传感重构效果,把流形学习的思想和方法与压缩传感相结合,提出了一种基于参数化流形学习的压缩传感重构方法。实验结果表明,提出的方法对自然图像进行重构取得了很好的效果,充分验证了基于参数化流形学习的压缩传感重构方法的有效性。

Compressive sensing reconstruction method based on parametric manifold learning

Compressive sensing is a new theory of information acquisition, which breaks through the traditional sampling theory. It combines data acquisition with data compression, and then recovers the original signal by reconstruction algorithms. In order to get the better effect of compressive sensing construction, this paper mainly combined the ideas and methods of manifold learning with compressive sensing, and then proposed a compressive sensing reconstruction method based on parametric manifold learning. The experimental results show that reconstruction of natural images has very good results with the method proposed in this paper. Therefore, it fully verifies the effectiveness of compressive sensing reconstruction method based on parametric manifold learning.