Presentation | 2013-03-11 A Study on Smoke Detection Method Based on LBP Featured and AdaBoost Yusuke IIDA, Hidenori MARUTA, Fujio KUROKAWA, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Smoke occurs simultaneously with a fire and we can observe it even if the fire is not in the field of view of the camera because it has features such as form change and spreading caused by its own growth or window. Therefore, smoke is a key feature to detect a fire. In this paper, we examine a smoke detection method using the image information aiming at the early detection of a fire. To detect smoke from image information which is acquired from a single camera, we use Local Binary Patterns (LBP) as an image feature. We compute the LBP histogram of each divided block using LBP and discriminate it as an input vector by AdaBoost with easy implement of an algorithm. In smoke detection, its results may not be robust and affected from environmental conditions such as background objects. Thus we adopt AdaBoost to obtain the robust detection results. We manually choose smoke and non-smoke data from the image sequences which have different environmental conditions for training of AdaBoost. Using trained classifier, the divided image block is classified whether it is smoke or non-smoke. In experiments with real world data, we examined our presented method can be robust for smoke detection. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Smoke detection / LBP / AdaBoost |
Paper # | IMQ2012-44,IE2012-148,MVE2012-105,WIT2012-54 |
Date of Issue |
Conference Information | |
Committee | MVE |
---|---|
Conference Date | 2013/3/4(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Media Experience and Virtual Environment (MVE) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | A Study on Smoke Detection Method Based on LBP Featured and AdaBoost |
Sub Title (in English) | |
Keyword(1) | Smoke detection |
Keyword(2) | LBP |
Keyword(3) | AdaBoost |
1st Author's Name | Yusuke IIDA |
1st Author's Affiliation | Graduate School of Engineering, Nagasaki University() |
2nd Author's Name | Hidenori MARUTA |
2nd Author's Affiliation | Graduate School of Engineering, Nagasaki University |
3rd Author's Name | Fujio KUROKAWA |
3rd Author's Affiliation | Graduate School of Engineering, Nagasaki University |
Date | 2013-03-11 |
Paper # | IMQ2012-44,IE2012-148,MVE2012-105,WIT2012-54 |
Volume (vol) | vol.112 |
Number (no) | 474 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |