1: \begin{abstract}

2: 

3: Many material and biological samples in scientific imaging 

4: are characterized by non-local repeating structures. 

5: These are studied using scanning electron microscopy and electron tomography.

6: Sparse sampling of individual pixels in a 2D image acquisition geometry, or sparse sampling of projection images with large tilt increments in a tomography experiment, can enable high speed data acquisition and minimize sample damage caused by the electron beam.  

7: 

8: In this paper, we present an algorithm for electron tomographic reconstruction 

9: and sparse image interpolation that exploits the non-local redundancy in images. 

10: We adapt a framework, termed plug-and-play (P\&P) priors, 

11: to solve these imaging problems in a regularized inversion setting. 

12: The power of the P\&P approach is that it allows a wide array of modern denoising algorithms to be used as a ``prior model'' for tomography and image interpolation.

13: We also present sufficient mathematical conditions that ensure convergence of the P\&P approach, and we use these insights to design a new non-local means denoising algorithm. 

14: Finally, we demonstrate that the algorithm produces higher quality reconstructions on both simulated and real electron microscope data, along with improved convergence properties compared to other methods.

15: 

16: \end{abstract}

17: