2021, Number 2
Influence of Interpolation on the Quality of Cell Microscopy Images
Coca RA, Lorenzo GJV
Language: Spanish
References: 12
Page:
PDF size: 604.48 Kb.
ABSTRACT
The present work has the purpose of analyzing, for the particular case of cell microscopy images of erythrocytes from human blood, to what extent the application of interpolation methods to improve the image resolution can influence the image quality and with which methods satisfactory results might be obtained. Three interpolation methods were applied for their comparison to the selected color images: cubic splines, bicubic and bilinear and their computational efficiency was also evaluated. Two resolution reduction factors (2 and 4) were used for rows and columns of the digital image. The measures used to assess the quality of the interpolated images were the signal-to-noise ratio and the mean square error, whose values were statistically processed using the Friedman and Wilcoxon tests, the latter as a post-hoc test. The results make it possible to recommend the bicubic interpolation method as the most favorable for this type of images since it was the one with the best performance among those used.REFERENCES
Rosenberger C, Chabrier S, Laurent H, Emile B. Unsupervised and supervised image segmentation evaluation. In: Advances in image and video segmentation. USA: IGI Global; 2006 [cited 19 Mar 2021]. 365-93. Available from: https://www.researchgate.net/profile/Christophe-Rosenberger/publication/257365161_Unsupervised_and_Supervised_Image_Segmentation_Evaluation/links/02e7e53597053b83f6000000/Unsupervised-and-Supervised-Image-Segmentation-Evaluation.pdf.