Delić, Marija, Ralević, Nebojša, Pap, Endre, Čomić, Lidija, Lukić, Tibor, Ćirović, Nataša, and Nedović, Ljubo
Problemi klasifikacije i segmentacije digitalnih slika su veomaaktuelni i zastupljeni u praksi. Potreba za modelima koji razmatrajuovu problematiku u poslednjih nekoliko decenija ubrzanim tempompoprima sve veći značaj i obim u svakodnevnom životu. Koriste se uračunarskoj grafici, prepoznavanju oblika, medicinskoj analizi slika,saobraćaju, analizi dokumenata, pokreta i izraza lica i sl.U okviru ove disertacije, predstavljeno istraživanje motivisano jeprimenama razvijenih modela u klasifikaciji i segmentacijidigitalnih slika. Istraživanje obuhvata dva segmenta. Ovi segmentipovezani su terminom neodređenosti, koji je uz upotrebu adekvatnogmatematičkog aparata (teorije fazi skupova), ugrađen u modele razvijeza primenu u obradi slike.Jedan pravac istraživanja baziran je na teoriji fazi skupova, t-normama, t-konormama, operatorima agregacije i agregiranimfunkcijama rastojanja. U okviru toga, istraživanje je sprovedeno sastruktuiranom matematičkom podlogom, izložene su osnovnedefinicije, teoreme, kao i osobine korištenih operatora, proširenisu teorijski koncepti t-normi i t-konormi. Definisani su novi tipovioperatora agregacije i njihovom primenom konstruisane su novefunkcije rastojanja, čija je upotreba diskutovana kroz uspešnost uprocesu segmentacije digitalnih slika.Drugi pravac istraživanja, izložen u ovoj disertaciji, obuhvata višeinženjerski pristup rešavanju problema klasifikacije teksturadigitalnih slika. U skladu sa tim, detaljno je analizirana idiskutovana klasa lokalnih binarnih deskriptora teksture.Inspirisana uspešnošću pomenute LBP klase deskriptora, uvedena jejedna nova podfamilija α-deskriptora teksture. Uvedeni modeldeskriptora formiran je na temeljima idejnih principa lokalnihbinarnih kodova i bazičnih pojmova iz teorije fazi skupova. Praktičnaupotreba i značaj predstavljenog modela demonstrirani su kroz veomauspešne procese klasifikacije na nekoliko javno dostupnih baza slika., Classification and segmentation problems of digital images is a very attractivetopic and has been making impact in many different applied disciplines. In thepast few decades, the demand for models that address these issues has beengaining momentum and applications in everyday life. These models are used incomputer graphics, shape recognition, medical image analysis, traffic, documentanalysis, facial movements and expressions, etc.The research within this doctoral dissertation was motivated by the application ofdeveloped methods in classification and segmentation tasks. The conductedresearch covered two segments, which were linked by the term of indeterminacy,with the usage of the theory of fuzzy sets, which is incorporated into methodsdeveloped for application in image processing.One direction of the research was founded on the theory of fuzzy sets, t-norms,t-conorms, aggregation operators, and aggregated distance functions. Within thisframework, the research was conducted with a structured mathematicalbackground. Firstly, basic definitions, theorems and characteristics of the usedoperators were presented, followed by the theoretical concepts of t-norms and tconormsthat were extended. New types of aggregation operators and distancefunctions were defined, and finally, their contribution in the digital imagesegmentation process was explored and discussed.The second direction of the research presented in this dissertation involved moreof an engineering-type of approach to solving the problem of the classification ofdigital image textures. To that end, a class of local binary texture descriptors(LBPs) was analyzed and discussed in detail. Inspired by the results of theabove-mentioned LBP descriptors, one new sub-family of the $\alpha$-descriptors was introduced by the author. The introduced descriptor model wasbased on the conceptual principles of LBPs and basic definitions from the fuzzyset theory. Its practical usage and importance were established and reflected invery successful classification results, achieved in the application on severalpublicly available image datasets.