基于改进StOMP算法图像压缩感知重构

发布时间：2018-01-01 01:33

本文关键词：基于改进StOMP算法图像压缩感知重构　出处：《计算机应用研究》2016年09期 　论文类型：期刊论文

【摘要】：在图像压缩感知重建中,针对重构效果和耗时不能兼得的问题进行深入研究。基于小波域稀疏,选用常规观测矩阵进行观测采样,通过对观测结果预定义滤波、选取信号硬阈值,引入共轭梯度下降算法,对分段正交匹配追踪(St OMP)重建算法进行改进。提出重建图像的边缘相似度概念,并对不同压缩比下的观测信号重建进行实验仿真。结果表明,相对于改进前St OMP算法,改进算法在迭代收敛时间较短的情况下,重构效果有所提升。在主观评价上,重建图像噪声点明显减少;客观评价上,PSNR值提高,达到了预期效果。该算法为高效解决压缩感知优化重建问题提供了参考。
[Abstract]:In the compressed sensing image reconstruction, to conduct a thorough study on the reconstruction effect and time have not. The wavelet domain based on sparse matrix, using routine observation sampling, using predefined observations of selected signal filtering, hard threshold algorithm, descending conjugate gradient, the piecewise orthogonal matching pursuit (St OMP) improved reconstruction the concept of similarity algorithm. The proposed edge image reconstruction, and the different compression ratio of the observed signal reconstruction simulation experiments. The results show that compared to the improved St OMP algorithm, the improved algorithm in convergence time is very short, the reconstruction effect improved. In the subjective evaluation, the noise image is obviously decreased; objective evaluation on the PSNR value increased, to achieve the desired effect. The algorithm is efficient to solve the optimization problem of compressed sensing reconstruction provides a reference.

【作者单位】：东北石油大学电气信息工程学院;
【基金】：黑龙江省自然科学基金资助项目(F201404)
【分类号】：TP391.41
【正文快照】： 0引言压缩感知理论是由Donoho和Candes[1~3]于2005年提出的一种以信号稀疏性为压缩先验条件的全新的信号采样理论。该理论表明,用远低于Nyquist采样定理要求的频率对信号进行采样也能实现信号的精确重构。压缩感知理论利用原始图像或信号的稀疏性先验知识,通过适当的稀疏优化

【相似文献】