dc.contributor.advisor | Thomassen, Asbjørn | |
dc.contributor.author | Støren, Henrik Hjorthen | |
dc.date.accessioned | 2017-03-13T07:37:57Z | |
dc.date.accessioned | 2017-03-13T07:37:57Z | |
dc.date.available | 2017-03-13T07:37:57Z | |
dc.date.available | 2017-03-13T07:37:57Z | |
dc.date.created | 2015-07-08 | |
dc.date.issued | 2015 | |
dc.identifier | ntnudaim:13899 | |
dc.identifier.uri | http://hdl.handle.net/11250/2353633 | |
dc.description.abstract | Hand gesture recognition is the task of having a machine recognize the hand gestures made by a human. In this thesis the main focus has been to research AI methods for gesture recognition. I investigate machine learning methods and image processing techniques to see if they are suited for hand gesture recognition with a color camera. I have used a PlayStation Eye camera, and written two different programs that use images captured by it to recognize and distinguish between 6 different static gestures. One of the programs uses only image processing techniques to recognize the gestures, while the other uses image processing to construct a feature vector, and then uses the KNN algorithm to predict the gesture in the image. The thesis is a proof of concept that will show the results of both programs and compare the two different approaches. I will also give a comparison between my approaches and what other researchers have done. | |
dc.language | eng | |
dc.publisher | NTNU | |
dc.subject | Datateknologi, Intelligente systemer | |
dc.title | Two-hand, camera-based gesture recognition for SoundDream | |
dc.type | Master thesis | |