Pamplona, Spain--Aránzazu Jurío-Munárriz, a graduate student in computer engineering at the NUP/UPNA-Public University of Navarre, has, for her PhD thesis, improved two of the most widespread algorithms used in digital image processing—thresholding and magnification.1
Thresholding
Image thresholding is used to resolve problems in areas such as remote sensing, where it is necessary to locate specific objects like rivers, forests, or crops in aerial images; the analysis of medical tests to locate different structures (organs, tumors, and so on), measuring the volumes of tissue and carrying out computer-guided surgery; and pattern recognition (fingerprints, car license plate numbers, etc.).
"Image thresholding separates out each of the objects that comprise the image," says Jurío-Munárriz. "To do this, each of the pixels is analyzed so that all the ones sharing the same features are considered to form part of the same object." In her PhD thesis, she presented two thresholding algorithms: the first adapted to working with fingerprint images, and the second geared toward brain images obtained by magnetic-resonance imaging (MRI) scans.
The first algorithm is for the NUP/UPNA's research group she belongs to, which is collaborating on a project to create an identification center by means of fingerprints that is capable of handling 40 million prints.
The aim of the second algorithm is to study the differences in the shapes or volumes of certain areas of the brain in patients who are suffering their first psychotic episodes. The researchers have come up with a method to be able to correctly separate out the area occupied by different brain structures in the image.
Magnification
The algorithm to magnify images stands out not only due to the quality obtained but also due to the time it takes to execute, which is 700 times less than other methods that obtain the same quality.
Jurío-Munárriz developed two new magnification methods, one for gray-scale images and the other for color images. The methods were developed to solve a problem of an infographics company. Starting with a 3D model, the company would generate various large images to show to its clients; generating them took more than 20 hours per image. The new algorithm allows images to be generated in a smaller size and then enlarged in a very short time while maintaining quality.
REFERENCE:
1. Humberto Bustince et al., European Journal of Operational Research, 225 - 3, p. 472 (2013).
