Statistical analysis of nanofiber mat AFM images by gray-scale-resolved Hurst exponent distributions

Author Błachowicz T.; Domino K.; Koruszowic M.; Grzybowski J.; Böhm T.; Ehrmann A.
Title Statistical analysis of nanofiber mat AFM images by gray-scale-resolved Hurst exponent distributions
Journal Applied Sciences
Year 2021
Status Published
Volume 11
Issue 2436
DOI https://doi.org/10.3390/app11052436
URL https://www.mdpi.com/2076-3417/11/5/2436/pdf
Abstract <p><span style="font-family:serif; font-size:14.1113px">Two-dimensional structures, either periodic </span><span style="font-family:serif; font-size:14.1113px">or random, can be classified by diverse </span><span style="font-family:serif; font-size:14.1113px">mathematical methods. Quantitative</span><span style="font-family:serif; font-size:14.1113px"> descriptions of such surfaces, however, are scarce since bijec-</span><span style="font-family:serif; font-size:14.1113px">tive definitions must be found to measure unique dependency between described structures and </span><span style="font-family:serif; font-size:14.1113px">the chosen quantitative parameters. To solve this </span><span style="font-family:serif; font-size:14.1113px">problem, we use statistical analysis of periodic </span><span style="font-family:serif; font-size:14.1113px">fibrous structures by Hurst exponent distribution</span><span style="font-family:serif; font-size:14.1113px">s. Although such a Hurst exponent approach was </span><span style="font-family:serif; font-size:14.1113px">suggested some years ago, the quantitative analysis of atomic force microscopy (AFM) images of </span><span style="font-family:serif; font-size:14.1113px">nanofiber mats in such a way was described only recently. In this paper, we discuss the influence of </span><span style="font-family:serif; font-size:14.1113px">typical AFM image post-processing steps on the gr</span><span style="font-family:serif; font-size:14.1113px">ay-scale-resolved Hurst exponent distribution. </span><span style="font-family:serif; font-size:14.1113px">Examples of these steps are polynomial background </span><span style="font-family:serif; font-size:14.1113px">subtraction, aligning rows, deleting horizontal </span><span style="font-family:serif; font-size:14.1113px">errors and sharpening. Our results show that whil</span><span style="font-family:serif; font-size:14.1113px">e characteristic features of these false-color </span><span style="font-family:serif; font-size:14.1113px">images may be shifted in terms of gray-channel and Hurst exponent, they can still be used to </span><span style="font-family:serif; font-size:14.1113px">identify AFM images and, in the next step, to qu</span><span style="font-family:serif; font-size:14.1113px">antitatively describe AFM images of nanofibrous </span><span style="font-family:serif; font-size:14.1113px">surfaces. Such a gray-channel approach can be regarded as a simple way to include some infor-</span><span style="font-family:serif; font-size:14.1113px">mation about the 3D structure of the image.</span></p>