Algorithm 3: Data Reconstruction Using Matching Pursuit Algorithm

Input: compressed data $y=r\times d$

Output: Reconstructed data $\overline{d}$

1:Initialize:

$\overline{d_0}$ = 0, $r_{error}=0$ and i = 0

2:iteration:

3:  i = i + 1

4:  $PN=\overline{d_{i-1}}+r^T\times r_{error}$

5:  $\overline{d_i}=argma{x}_{i=\left\{1,2,..N\right\}}\left|PN\right|$

6:Compute residual error

$r_{error}=y-r\times \overline{d_i}$

7:If there is no error in $r_{error}$then

8:return d

9:End

10:End