L'objectiu d'aquest TFC es la implementacio d'eines software per a poder avaluar i comparar diferents algoritmes d'analisi PIV (Particle Image Velocimetry). Primerament, es defineix i explica la tecnica PIV. El PIV es un metode optic per a mesurar el camp de velocitats en un flux. Es basa en la captura d'imatges d'un flux amb particules de seguiment afegides que, mitjancant un processat digital i analisis de les mateixes, determina el camp de velocitat instantania del fluid que circula. Es important, abans de voler avaluar els diferents tipus d'algorismes, entendre com funcionen i saber d'entrada les seves limitacions teoriques. En aquest sentit, s'hna analitzat quines son les diferents configuracions d'interrogacio iquins avantatges i inconvenients ofereixen. El software que s'ha analitzat es el GPIV. Aquest es un software de codi lliure creat l'any 2003, i des de llavors ha sofert importants modificacions en els seus algorismes. Els tests han analitzat, basicament, la resposta espectral dels resultats, aixi com tambe la distribucio dels valors decimals de velocitat, el que es coneix com a efecte peak-locking. Un cop obtinguts els resultats, s'han analitzat i comparat amb els resultats esperats i amb resultats obtinguts utilitzant diferents parametres d'interrogacio. L'informe de resultats del Piv challenge ha estat realment util i instructiu per a la realitzacio d'aquest TFC. I, finalment, es conclouen quins son els aspectes que han millorat en el GPIV, en quins es pot millorar mes, i es proposen altres analisis que podrien ser utils a l'hora d'avaluar futures modificacions.

The aim of this project is the implementation of a set of software tools to evaluate and compare different advanced PIV interrogation algorithms. Firstly, Particle Image Velocimetry (PIV) technique will be explained. PIV is an optical method for flow visualization. When using PIV, a fluid is seeded with tracer particles and subsequent pictures (mostly two) of the flow are taken with a digital camera. After some digital processing we are able to compute the value of the average displacement of the seeding particles within a small volume or area of the flow. The estimator of the particle displacement value is used to calculate the local speed and direction (the velocity field) of the flow being studied. From a single image, several estimators of particle displacements can be calculated at a grid, resulting into a two-dimensional velocity field. Before testing the advanced PIV algorithms, it's important to know how they work, and what are supposed to be their main advantages and disadvantages. For this purpose, we'll analyze the main advanced interrogation methods. The software that will be tested is called GPIV. It's an open-source software released on 2003. Since then, their interrogation algorithms have been substantially upgraded. The created tools test, the spectral resolution of the interrogation algorithms, and the distribution of particle displacements at subpixel level, in order to examine the so-called peak-locking effect. The obtained results will be compared to the theoretically expected, as well as to previous results from GPIV algorithms, obtained when GPIV took part as a contributor in a international PIV challenge on 2005. The main results of that challenge have been very useful and instructive to complete this project. Finally, we'll conclude naming what are the aspects of GPIV which have been improved, in which of them there is still room for improvements, and we'll also suggest some tests that could be useful when evaluating further modifications or upgrades on GPIV software.