Automatické určení tvaru a velikosti nanočástic pro biomedicínské aplikace

Abstract

Nanotechnology is widely used in biomedical research. For its usage it is important to know sizes and shapes of particles. These parameters can be measured by transmission electron microscopy (TEM), but analysis of images from transmission electron microscope is complicated and tedious. It would be easier if method for automatic evaluation of these parameters would exist. The goal of this project is to create program for automatic determination of sizes and shape of nanoparticles and nanorods from TEM images. TEM samples of nanoparticles and nanorods of various sizes were used and TEM images were acquired. Image analysis methods were used for image adjustement and two image segmentation methods were used. First method is watershed transform and second method is circle Hough transform. Python program was created and tested on various images of nanoparticles and nanorods. Sizes and shapes of nanoparticles and nanords were calculated and histograms of sizes were plotted.

Nanotechnology is widely used in biomedical research. For its usage it is important to know sizes and shapes of particles. These parameters can be measured by transmission electron microscopy (TEM), but analysis of images from transmission electron microscope is complicated and tedious. It would be easier if method for automatic evaluation of these parameters would exist. The goal of this project is to create program for automatic determination of sizes and shape of nanoparticles and nanorods from TEM images. TEM samples of nanoparticles and nanorods of various sizes were used and TEM images were acquired. Image analysis methods were used for image adjustement and two image segmentation methods were used. First method is watershed transform and second method is circle Hough transform. Python program was created and tested on various images of nanoparticles and nanorods. Sizes and shapes of nanoparticles and nanords were calculated and histograms of sizes were plotted.