undergraduate thesis
Making connect-the-dots game by web cam motion detection

Zoran Horvat (2015)
Sveučilište Josipa Jurja Strossmayera u Osijeku
Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek
Zavod za programsko inženjerstvo
Katedra za vizualno računarstvo
Metadata
TitleStvaranje igre "spoji ročkice" putem detekcije pokreta web kamerom
AuthorZoran Horvat
Mentor(s)Časlav Livada (thesis advisor)
Abstract
Zadani problem završnog rada bio je izraditi računalni program koji će pomoću kamere povezane s računalom detektirati zadani objekt na kameri, pratiti ga i pomoću kretnji objekta periodički iscrtavati segmente (točke) za podlogu za igru „Spoji točkice“ na bijeloj površini. Odabrani jezik za pisanje skripte koja će raditi traženi posao je skriptni jezik Python. Kao dodatak Pythonu, korištena je biblioteka vezana uz računalni vid i strojno učenje OpenCV. Označavanje željenog objekta radi se klikom i potezom miša, a praćenje se odvija na temelju usporedbe histograma i pomoću algoritma meanshift, koji predstavlja konstantno traženje najveće gustoće zadanih piksela na slici s kamere. Digitalna slika se pretvara iz RGB sustava boja u HSV. Rješenje je dano implementacijom ručke za praćenje pokreta miša, primjenom algoritma meanshift na svakoj slici s ulaza kamere te projekcijom pokreta objekta u stvarnom vremenu na bijelu podlogu i periodičko crtanjenumeriranih točaka kao segmenata podloge za igru. Također je na jednostavan način spriječeno preklapanje točaka provjeravanjem okoline trenutne pozicije objekta preslikane na bijelu podlogu. Konačna verzija skripte uspješno izvršava postavljen zadatak te na izlazu daje podlogu za igru generiranu pomoću detekcije pokreta u stvarnom vremenu.
Keywordscomputer vision object detection motion tracking digital image meanshift algorithm histogram pixel density color space drawing.
Parallel title (English)Making connect-the-dots game by web cam motion detection
Committee MembersČaslav Livada (committee chairperson)
Hrvoje Glavaš (committee member)
Zoran Balkić (committee member)
GranterSveučilište Josipa Jurja Strossmayera u Osijeku
Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek
Lower level organizational unitsZavod za programsko inženjerstvo
Katedra za vizualno računarstvo
PlaceOsijek
StateCroatia
Scientific field, discipline, subdisciplineTECHNICAL SCIENCES
Computing
Program Engineering
Study programme typeuniversity
Study levelundergraduate
Study programmeUndergraduate University Study Programme in Computer Engineering
Academic title abbreviationuniv.bacc.ing.comp.
Genreundergraduate thesis
Language Croatian
Defense date2015-07-09
Parallel abstract (English)
The project assignment was to design a computer application which should capture an object on webcam, track its movement and repeatedly draw dots for the „Connect the dots“ game on a white surface. The chosen language to engage the project with was a scripting language called Python. As a Python extension, a premade library called OpenCV was used. The library is in charge of providing an interface to work with computer vision and machine learning. The object selection on live camera feed is done by clicking and dragging the mouse over the object that should be tracked. The tracking is based on histogram comparison and meanshift algorithm which represents an uninterrupted search for the largest density of initially selected pixels. The digital image is converted from RGB to HSV color space. The solution is given as a combination of a mouse handle implementation, application of meanshift algorithm on each camera frame and projection of the real-time object movement on a separate white background which results in periodical numerated black dot drawing on the said background. The dot overlapping prevention was also included with simple environment checks around the current object position on white surface. The final script version successfully finishes the project assignment by outputting the game base generated using real-time object tracking.
Parallel keywords (Croatian)računalni vid detekcija objekta praćenje pokreta digitalna slika meanshift algoritam histogram gustoća piksela sustav prikaza boja crtanje.
Resource typetext
Access conditionOpen access
Terms of usehttp://rightsstatements.org/vocab/InC/1.0/
URN:NBNhttps://urn.nsk.hr/urn:nbn:hr:200:820714
CommitterAnka Ovničević