General Image Fiber Tool A concept for automated evaluation of fiber diameters in SEM images

Abstract Fiber imaging is becoming increasingly important in various fields. The current standard method of quality assurance is the manual quantification of fiber diameters in Scanning Electron Microscope (SEM) images, which is time consuming and introduces bias. However, due to partly insufficient reliability of automated software, improved concepts are required. We introduce an innovative routine for computerized evaluation of fiber diameters in SEM images with improved accuracy. This General Image Fiber Tool (GIFT) automatically calculates the average fiber diameter and standard deviation by statistical analysis. In a comparative study, GIFT was benchmarked to an existing popular fiber analysis tool, DiameterJ, and manual quantification. GIFT has detected fiber diameters with improved accuracy in numerous SEM images, and has been shown to be a remarkably sound method when challenged with difficult image types. The work presented here validates the GIFT concept and demonstrates its potential for applications in biomedicine and various other fields.