eMonocot Cyperaceae

an authoritative resource for Cyperaceae data worldwide, integrating global and regional perspectives

The start-to-end chemometric image processing of 2D thin-layer videoscans

Publication Type:Journal Article
Year of Publication:2011
Authors:Komsta, L., Ciesla, L., Bogucka-Kocka, A., Jozefczyk, A., Kryszen, J., Waksmundzka-Hajnos M.
Journal:Journal of Chromatography A
Keywords:2d-tlc, algorithms, Carex, carex spp., coupled device noise, Cyperaceae, fingerprinting, gel-electrophoresis images, image processing, optimization, planar chromatography, separation, tlc

The purpose of the research was to recommend a unified procedure of image preprocessing of 20 thin layer videoscans for further supervised or unsupervised chemometric analysis. All work was done with open source software. The videoscans saved as JPG files underwent the following procedures: denoising using a median filter, baseline removal with the rollerball algorithm and nonlinear warping using spline functions. The application of the proposed procedure enabled filtration of random difference between images (background intensity changes and spatial differences of the spots location). After the preprocessing only spot intensities have an influence on the performed PCA or other techniques. The proposed technique was successfully applied to recognize the differences between three Carex species from the 20 videoscans of the extracts. The proposed solution may be of value for the any chemometric task - both unsupervised and supervised. (C) 2010 Elsevier B.V. All rights reserved.

Scratchpads developed and conceived by (alphabetical): Ed Baker, Katherine Bouton Alice Heaton Dimitris Koureas, Laurence Livermore, Dave Roberts, Simon Rycroft, Ben Scott, Vince Smith