Primerjava rezultatov obdelave termogramov

PAVLICA, VITO (author), Pušnik, Igor (mentor)

Že nekaj časa se na tematiko raziskav pojavlja veliko dvomov ter razprav o njihovi verodostojnosti. Uporabljene metode in izvedbe so včasih veliko bolj usmerjene k temu, da se raziskava izvede, kot pa da je izvedena pravilno. Na temo obnovljivosti je ogromno zapisov in kritik, saj kot poročajo mnogi viri, raziskav za potrditev obnovljivosti redki želijo financirati. Večina si želi zgolj nove raziskave, ki bi pripeljale do novih odkritij in nadaljnjih raziskovanj. Kar je poleg tega tudi zaskrbljujoče, je dejstvo, da veliko število raziskav ne omogoča obnovljivosti, kar nakazuje na pristranskost, česar v raziskovanju in znanosti ne bi smelo biti. V diplomski nalogi je pokrita zgornja tematika. Primerjava rezultatov obdelave termogramov pri različnih obdelovalcih nam daje vpogled v raziskovanje obnovljivosti. Po obdelavi raziskovalcev sem termograme obdelal še sam v programu FLIR ResearchIR, kjer sem poizkušal obrisati enako področje zanimanja (ROI), kot so ga raziskovalci pred mano. V obeh primerih je le-to bilo precej enostavno, saj so bili podatki definirani precej natančno in nisem imel večjih problemov z razumevanjem tematike ter področij zanimanja. Meritve so bile izvedene s kamero FLIR T650sc in sicer na dlaneh telovadcev ter v okolici vratnega predela v drugi raziskavi. Pri telovadcih so bila polja zanimanja manjša, vsaka dlan je bila razdeljena na 9 ROI-jev. V medicinski raziskavi pa je bilo območje zanimanja zgolj eno, okolica vratnega vretena, kjer so bili subjekti elektro stimulirani. Moje diplomsko delo bo raziskovalcem služilo kot ovrednotenje njihovih nalog, saj so rezultati delni pokazatelj kakovosti oziroma negotovosti njihovih del. Služi pa tudi kot referenca mentorju, saj je na tak način prikazano ali so bile raziskave primerno zastavljene ter izvedene. Izkazalo se je, da je obnovljivost pri obeh raziskavah mogoča, vendar v veliki meri odvisna že od kvalitete raziskave same. Če so raziskava in posnetki izpeljani zelo kvalitetno, je mogoče dosegati izjemno obnovljivost. V nasprotnem primeru pride do večjih odstopanj. Raziskavi, ki sem jih analiziral, sta zelo pomembni za svoja področja, saj bosta uporabljeni kot osnova za nadaljnjo raziskovanje področij, na katere se nanašata.

• Language: Slovenian
• Keywords: termografija, termogrami, obnovljivost, merilna negotovost, področje zanimanja
• Work type: Bachelor thesis/paper
• Organization: FE - Faculty of Electrical Engineering
• Year: 2020
• PID: 20.500.12556/RUL-119134
• Publication date in RUL: 03.09.2020

Secondary language

• Language: English
• Title: Comparison of results in analysis of thermograms

Abstract

There has been a lot of doubt about research for some time now. The methods and implementations used are sometimes much more focused on getting the research done than on doing it right. There are huge records and criticisms on the topic of reproducibility, because as many sources report, only a few wants to fund reproducibility research. Majority just wants new research that would lead to new discoveries and further research. What is also worrying is the fact that a large number of studies do not allow for reproducibility - which indicates a bias, which should not be the case in research and science. My bachelor thesis covers the above topic. A comparison of the results of thermogram processing of different analysts gives us an insight into the research of reproducibility. After processing by the researchers, I also processed the thermograms myself in the FLIR ResearchIR program, where I tried to outline the same region of interest as the researchers before me. In both cases this was fairly straightforward, as the data was defined quite accurately and I had no major problems understanding the topic and areas of interest. Measurements were performed with a FLIR T650sc camera on the palms of gymnasts and around the neck in the other study. In gymnasts, the fields of interest were smaller - each palm was divided into 9 ROIs. In medical research, however, the region of interest was only one - the neck area, where subjects were electrically stimulated. My dissertation will serve researchers as an evaluation of their assignments, as the results are a partial indicator of the uncertainty of their work. It also serves as a reference to the mentor for confirmation whether the research has been properly set up and carried out. Reproducibility has been shown to be possible in both studies, but largely depends on the quality of the study itself. If the research and recordings are carried out with high quality, exceptional reproducibility can be achieved. Otherwise, major deviations occur. The research I have analyzed is very important for their fields, as they will be used as a basis for further research in the areas they are working.

• Keywords: thermography, thermograms, reproducibility, measurement uncertainty, region of interest