Abstract
Video tools developed today, teachers can record lecture videos and upload these lecture videos to e-learning system themselves. However, some students may only do not understand some fragments but they have to waste unnecessary time download entire video, and therefore video scene segmentation is relative importance. In addition, in traditional teaching model, students must listen and transcribe content on blackboard, when the lecture's write speed is too fast, students are very difficult to pay attention in class and easy to transcription errors. If we can record the lecture video and automatically extract content on the blackboard, students not only enable index videos easily, but also not easily transcribe the wrong notes. Hence, this paper presents an intelligent assistance system for lecture videos. We use K-mean Segmentation to extract blackboard area and then only update this area to avoid lecturer's body cover the content on blackboard. Then we use adaptive threshold to extract chalk and design a method to reduce noise. After that we detecting the lecture videos split timing with statistics chalk pixel count and analysis the variation. Finally, we further interception the most complete content as lecture note to facilitate the students to search for video clips.
| Original language | English |
|---|---|
| Title of host publication | Proceedings - 2014 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014 |
| Editors | Junzo Watada, Akinori Ito, Chien-Ming Chen, Jeng-Shyang Pan, Han-Chieh Chao |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 357-360 |
| Number of pages | 4 |
| ISBN (Electronic) | 9781479953905 |
| DOIs | |
| Publication status | Published - 2014 Dec 24 |
| Event | 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014 - Kitakyushu, Japan Duration: 2014 Aug 27 → 2014 Aug 29 |
Publication series
| Name | Proceedings - 2014 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014 |
|---|
Other
| Other | 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014 |
|---|---|
| Country | Japan |
| City | Kitakyushu |
| Period | 14/8/27 → 14/8/29 |
Fingerprint
Keywords
- Image Processing
- Notes Extraction
- Video Segmentation
- Video Summarization
ASJC Scopus subject areas
- Information Systems
- Artificial Intelligence
- Signal Processing
Cite this
Robust handwriting extraction and lecture video summarization. / Yeh, Fu Hao; Lee, Greg C; Chen, Ying Ju; Liao, Chien Hsing.
Proceedings - 2014 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014. ed. / Junzo Watada; Akinori Ito; Chien-Ming Chen; Jeng-Shyang Pan; Han-Chieh Chao. Institute of Electrical and Electronics Engineers Inc., 2014. p. 357-360 6998341 (Proceedings - 2014 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Robust handwriting extraction and lecture video summarization
AU - Yeh, Fu Hao
AU - Lee, Greg C
AU - Chen, Ying Ju
AU - Liao, Chien Hsing
PY - 2014/12/24
Y1 - 2014/12/24
N2 - Video tools developed today, teachers can record lecture videos and upload these lecture videos to e-learning system themselves. However, some students may only do not understand some fragments but they have to waste unnecessary time download entire video, and therefore video scene segmentation is relative importance. In addition, in traditional teaching model, students must listen and transcribe content on blackboard, when the lecture's write speed is too fast, students are very difficult to pay attention in class and easy to transcription errors. If we can record the lecture video and automatically extract content on the blackboard, students not only enable index videos easily, but also not easily transcribe the wrong notes. Hence, this paper presents an intelligent assistance system for lecture videos. We use K-mean Segmentation to extract blackboard area and then only update this area to avoid lecturer's body cover the content on blackboard. Then we use adaptive threshold to extract chalk and design a method to reduce noise. After that we detecting the lecture videos split timing with statistics chalk pixel count and analysis the variation. Finally, we further interception the most complete content as lecture note to facilitate the students to search for video clips.
AB - Video tools developed today, teachers can record lecture videos and upload these lecture videos to e-learning system themselves. However, some students may only do not understand some fragments but they have to waste unnecessary time download entire video, and therefore video scene segmentation is relative importance. In addition, in traditional teaching model, students must listen and transcribe content on blackboard, when the lecture's write speed is too fast, students are very difficult to pay attention in class and easy to transcription errors. If we can record the lecture video and automatically extract content on the blackboard, students not only enable index videos easily, but also not easily transcribe the wrong notes. Hence, this paper presents an intelligent assistance system for lecture videos. We use K-mean Segmentation to extract blackboard area and then only update this area to avoid lecturer's body cover the content on blackboard. Then we use adaptive threshold to extract chalk and design a method to reduce noise. After that we detecting the lecture videos split timing with statistics chalk pixel count and analysis the variation. Finally, we further interception the most complete content as lecture note to facilitate the students to search for video clips.
KW - Image Processing
KW - Notes Extraction
KW - Video Segmentation
KW - Video Summarization
UR - http://www.scopus.com/inward/record.url?scp=84921628202&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84921628202&partnerID=8YFLogxK
U2 - 10.1109/IIH-MSP.2014.95
DO - 10.1109/IIH-MSP.2014.95
M3 - Conference contribution
AN - SCOPUS:84921628202
T3 - Proceedings - 2014 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014
SP - 357
EP - 360
BT - Proceedings - 2014 10th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2014
A2 - Watada, Junzo
A2 - Ito, Akinori
A2 - Chen, Chien-Ming
A2 - Pan, Jeng-Shyang
A2 - Chao, Han-Chieh
PB - Institute of Electrical and Electronics Engineers Inc.
ER -