Extracting Indicated Objects in Lecture Materials during a Lecture

Takafumi MARUTANI  Satoshi NISHIGUCHI  Koh KAKUSHO  Michihiko MINOH  

D - Abstracts of IEICE TRANSACTIONS on Information and Systems (Japanese Edition)   Vol.J90-D   No.5   pp.1238-1248
Publication Date: 2007/05/01
Online ISSN: 1881-0225
Print ISSN: 1880-4535
Type of Manuscript: PAPER
lecture contents,  indicated object,  deictic gesture recognition,  indicated region,  

Full Text(in Japanese): PDF(643.8KB)
>>Buy this Article

In this paper, we propose a method to extract the indicated objects in computer presentations indicated by lecturer's deictic gesture during a lecture, because the information of objects indicated by lecturer's deictic gesture is useful for following applications; automatic shooting a lecturer, emphasizing objects in lecture materials, indexing of lecture contents, and so on. Originally, indicated objects are determined by both the edge position of pointing devices and the kind of deictic gesture (Pointing, Highlighting, Outlining, Emphasizing). Therefore, for extracting indicated objects, it is necessary to classify the kind of lecturer's deictic gesture. We propose the method for classifying the deictic gestures using new feature, lecturer's position, lecturer's orientation, and the kind of the lecture material's region, in addition to the orbit of indicated points. We verified the effectiveness of our method by the success rate of extraction.