Analysis of Nanopore Structure Images Using MATLAB Software

The importance of nanopores increases with time due to their application. For instance,
nanopores may be used to sense molecules like DNA and RNA, single proteins, etc. Sequencing by
nanopore has also a possibility to be a direct, fast, and inexpensive DNA sequencing tool. Diameters of
nanopores are the main keys for mentioned sensing processes. Three segmenting methods used in this
study namely Thresholding, Gaussian Mixture Model-Expectation Maximization (GMM-EM) and
Hidden Markov Random Field-Expectation Maximization (HMRF-EM). These methods applied on
three SEM nanopore images after enhancing them through obtaining optimum parameters of CLAHE
contrast-enhanced method to give high PSNR. The results of the Rand index and time of running code
show that the HMRF-EM is better than GMM-EM. Hence, their segmented images are used to find out
nanopore parameters including total counting pores, diameter, and porosity. The results of porosity
were in good agreement with former investigations. Consequently, the HMRF-EM segmenting technique
with procedures utilized in this study using image processing for finding porosity gives promising results
among other examined methods.