2015
Conference article  Open Access

PRIAR using a graph segmentation method

Righi M., D'Acunto M., Salvetti O.

Pattern Recognition  Image Analysis  Image Segmentation  Boundary Detection  Graph Partitioning Algorithm 

Recently, we have suggested a simple and general-purpose method able to combine high-resolution analysis with the classification and identification of components of microscopy imaging. The method named PRIAR (Pattern Recognition Image Augumented Resolution) is a tool developed by the authors that gives the possibility to enhance spatial and photometric resolution of low-res images. The implemented algorithm follows the scheme: 1) image classification; 2) blind super-resolution on single frame; 3) pattern-analysis; 4) reconstruction of the discovered pattern. In this paper, we suggest some improvements of the PRIAR algorithm, in particular, the definition of a segmentation method which is based on homomorphism between a processed image and a graph describing the image itself, able to identify object of interest in complex patterns. The case study is the identification of organs inside biological cells acquired with Atomic Force Microscopy Technique.

Source: IMTA 2015 - 5th International Workshop on Image Mining. Theory and Applications, pp. 46–51, Berlin, Germany, 11-14 March 2015

Publisher: SCITEPRESS - Science and Technology Publications, digital library, USA



Back to previous page
CNR ExploRA

Bibliographic record

ISTI Repository

Postprint version Open Access

Also available from

www.scitepress.orgRestricted

BibTeX entry
@inproceedings{oai:it.cnr:prodotti:327845,
	title = {PRIAR using a graph segmentation method},
	author = {Righi M. and D'Acunto M. and Salvetti O.},
	publisher = {SCITEPRESS - Science and Technology Publications, digital library, USA},
	booktitle = {IMTA 2015 - 5th International Workshop on Image Mining. Theory and Applications, pp. 46–51, Berlin, Germany, 11-14 March 2015},
	year = {2015}
}