ChildCrowds: A Computer Vision based Crowd Gaming framework for children

ChildCrowds: Et computer Vision baseret Crowd Gaming framework for børn

Hansen, Tommy Paaske

4. term

Medialogy, Master

2013

2013-08-30

108

Dette projekt lægger grunden til et kamerabaseret crowd-gaming‑framework for børn, der gør det muligt for grupper at lege og spille sammen. Arbejdet omfatter en gennemgang af teorier om masse- og gruppeadfærd, børns lege og legens udviklingsstadier, definitioner af spil og genrer samt en omfattende gennemgang af eksisterende projekter inden for crowd-interaktion. Derudover analyseres computersynsteknikker, mulige crowd‑gamingaktiviteter og metoder til at teste med en målgruppe af børn. På baggrund af analysen designede vi et framework, der bruger et modificeret PlayStation 3-kamera og farvesegmentering til at identificere selvlysende, kemiske lysstave i videostrømmen. Systemet sporer de gennemsnitlige positioner af grønne og blå lysstave, som børnene holder, og tre forskellige måder at interagere på blev udviklet ud fra disse positioner. Vi designede også tre spil for at evaluere frameworket gennem disse interaktionsmåder. Frameworket blev afprøvet og evalueret gennem observation af tre spilsessioner, spørgeskemaer med rangskalaer samt et fokusgruppeinterview med syv børn fra en fritidsklub.

This project lays the groundwork for a camera‑based crowd gaming framework for children, enabling groups to play and collaborate together. The work includes reviewing theories of crowd and group behavior, children’s play and its developmental stages, definitions of games and genres, and an extensive state‑of‑the‑art survey of existing crowd interaction projects. It also examines computer vision techniques, possible crowd gaming activities, and methods for testing with a child target group. Based on this analysis, we designed a framework that uses a modified PlayStation 3 camera and color segmentation to identify chemical glow sticks in the video feed. The system tracks the average positions of green and blue glow sticks held by players, and three different ways of interacting were built around those positions. We then created three games to evaluate the framework via these interaction modes. The framework was tested and evaluated through observations of three gaming sessions, rating‑scale questionnaires, and a focus group interview with seven children from an after‑school center.

[This abstract was generated with the help of AI]

Crowdgaming ; Crowd ; Unity3d ; Children ; Kids ; OpenCvSharp ; Computer Vision ; Image Processing ; Play ; Collaborative play

Download
View record in AAU Student Projects