Applying Principal Component Analysis for Macromolecular Objects Diffraction Images Sorting

In this paper we present a method for fast detection of images containing diffraction patterns of specific macromolecular objects. Our method is based on a principal component analysis, a popular method used in a various areas to analyse multi-dimensional data like image classification, noise reduction, video indexing and etc. In our paper we demonstrate that our method is efficient for diffraction images classification for various macromolecular structures containing collagen. Diffraction data was gathered from Kurchatov synchrotron radiation source.