Nanoparticles diameter characterization using image analysis methodology

Methods for automatic nanoparticles detection, shape analysis and size estimation have obtained importance to provide quantitative support that gives significant information for evaluation in Materials Science. This paper presents a methodology for automatic nanoparticles diameter characterization by image analysis. A nanoparticles images database was obtained by means of an electron microscope in the Autonomous University of Baja California (UABC) and other images were obtained from nanoComposix web page. Firstly, an adaptive Wiener Filtering Method (AWF) was carried out to remove noise and Contrast Limited Adaptive Histogram Equalization Method (CLAHE) as enhancement step. Later two segmentation methods were compared: K-means Clustering (KMC) and Adaptive Thresholding (AT), both with MATLAB programming environment. Results showed that KMC performed better than AT i.e., a KMC mean square error 41.98 vs AT MSE of 172.11.