CAJ | 학술논문

本文利用MATLAB平台设计了基于压缩感知的图像重构GUI系统，针对不同的变换基、测量矩阵、重构算法对图像的压缩采样及重构进行了比较和讨论，其中变换基包括小波变换基和离散余弦变换基；测量矩阵包括随机矩阵如高斯矩阵和贝努利矩阵，以及确定矩阵如哈达玛矩阵和托普利兹矩阵；重构算法包括正交匹配追踪算法和压缩采样匹配追踪算法。并通过峰值信噪比PSNR衡量了图像重构的性能。该GUI系统能够为基于压缩感知的图像重构提供直观的实现平台。
본문이용MATLAB평태설계료기우압축감지적도상중구GUI계통，침대불동적변환기、측량구진、중구산법대도상적압축채양급중구진행료비교화토론，기중변환기포괄소파변환기화리산여현변환기；측량구진포괄수궤구진여고사구진화패노리구진，이급학정구진여합체마구진화탁보리자구진；중구산법포괄정교필배추종산법화압축채양필배추종산법。병통과봉치신조비PSNR형량료도상중구적성능。해GUI계통능구위기우압축감지적도상중구제공직관적실현평태。
In this paper ,an image reconstruction GUI system based on compressive sensing is designed by MATLAB platform ,and the image reconstruction is compared and discussed using different transformation bases including the wavelet transform and the discrete cosine transform ,measurement matrices including random matrices such as Gaussian matrix and Bernoulli matrix , deterministic matrices including Hadamard matrix and Toeplitz matrix , as well as reconstruction algorithms including orthogonal matching pursuit algorithm and compressive sampling matching pursuit . Moreover , the image reconstruction performance is evaluated by using the peak signal‐to‐noise ratio (PSNR) .The proposed GUI system can provide a direct platform for the realization of the image reconstruction based on compressive sensing .