Prikaz interaktivnih 3D modelov v obogateni resničnosti na mobilnih napravah

Magistrsko delo se navezuje na dve raziskovalni področji, in sicer 3D-računalniško grafiko in obogateno resničnost s kombinacijo njunega prikazovanja na mobilnih napravah. Naloga predstavlja izdelavo treh različnih interaktivnih 3D-modelov, ki temeljijo na realistično risani podobi domačih živali in jih je mogoče prikazati s pomočjo obogatene resničnosti na mobilnih napravah. Živali se prikazujejo v izdelani aplikaciji Obogatene živali z osnovnim uporabniškim vmesnikom. Prikazovanje temelji na načelu vzorčne metode, kjer govorimo o zaznavanju med mobilno napravo in obogatenim papirjem. Slednje predstavlja slikovno tarčo v obliki natisnjene interaktivne kartice. Ko kamera mobilne naprave prepozna tarčo, prikaže izbrano žival na mobilnem zaslonu, kjer dobimo rezultat nadgradnje resničnega sveta s 3D-liki. Ob prikazu 3D-lika na zaslonu je s pomočjo uporabniškega vmesnika možno vsako posamezno žival spoznati še v okviru treh različnih animiranih gibanj. Prvi del magistrskega dela je bil izveden v programu Blender, kjer smo ustvarili vse tri živalske 3D-like. Najprej smo morali živali zmodelirati v prepoznavno 3D-obliko, ki smo jim v nadaljevanju prilepili teksture. Sledila je postavitev sistema kosti in animacijskih kontrol, na temelju katerega smo like animirali. Po tem koraku smo delo prenesli v program Unity. V njem smo pripravili platformo, ki omogoča prikazovanje živalskih likov v načinu obogatene resničnosti. Končni rezultati celotnega dela so ustrezno izdelani živalski liki v obliki animiranih 3D-modelov, ki jih je mogoče prikazati v načinu obogatene resničnosti na mobilnih napravah z uporabo interaktivnih kartic. Izbrani parametri testiranja so pokazali, da prihaja do določenih razlik pri predvajanju med obema mobilnima napravama glede na izbrano stopnjo subdivizije 3D-lika. Pri zaznavanju na temelju svetlosti, oddaljenosti in naklona med slikovno tarčo in mobilno napravo je najboljša uporabniška izkušnja dosežena, če je slikovna tarča zajeta iz razdalje 15–20 cm in iz pogleda ptičje perspektive ob dobrih svetlobnih pogojih. This master's thesis is related to two research areas, namely 3D computer graphics and augmented reality, with a combination of their presentation on mobile devices. The thesis presents the creation of three different interactive 3D models based on a realistically drawn image of domestic animals, which can be displayed on mobile devices using augmented reality. The textured animals' models are displayed in an application called Augmented Animals (sl. Obogatene živali) with a simple user interface. The usability of the application is demonstrated by its detection of the image target, i.e. a printed interactive card, which proves the interaction between the mobile device and the augmented paper. When the mobile camera recognizes the target, it displays the selected animal on the screen. The result is an enhancement of the real environment with animated 3D characters. By displaying a 3D character on the screen and interacting with the user interface, the presentation of each animal in three different animated movements is enabled. The first part of this thesis was carried out using the Blender program, where we created all three 3D animal characters. First, we had to model all of the animals from their initial templates into a recognizable 3D shape, to which we then pasted the textures. This was followed by the construction of a system of bones and animation controls, based on which we were able to make animal animations. After this step, we transferred the project to the Unity program, where we constructed an application that is able to show characters in augmented reality. The final result of the entire process are adequately developed animal characters in form of animated 3D models, which can be displayed in the augmented reality mode on mobile devices using interactive cards. The selected testing parameters showed that there are certain differences in presentation between the two mobile devices depending on the selected 3D character subdivision level. However, in terms of detection based on light conditions, distance and inclination between the image target and the mobile device, the best user experience is achieved if the image target is captured from a distance of 15–20 cm, on a bird's eye view and under good lighting conditions.