Flame Feature Model Development and Its Application to Flame Detection
| Autor: | Jian-Wen Peng, Wen-Bing Horng, Chien-Yuan Peng, Tai-Fang Lu |
|---|---|
| Rok vydání: | 2006 |
| Předmět: |
Masking (art)
business.industry Computer science Flame detection Feature extraction ComputerApplications_COMPUTERSINOTHERSYSTEMS Image segmentation Frame rate Combustion Linear discriminant analysis GeneralLiterature_MISCELLANEOUS Computer vision Artificial intelligence Spurious relationship business ComputingMethodologies_COMPUTERGRAPHICS |
| Zdroj: | ICICIC (1) |
| DOI: | 10.1109/icicic.2006.82 |
| Popis: | In this study, both linear discriminant technique and logistic regression method are considered and compared to develop a flame feature model. Based on the comparison, the logistic regression method is chosen and plugged into a flame detection system for segmenting the fire regions. Some spurious fire-like regions are then removed by the image difference method and the invented color masking technique. Finally, a simple method is devised to estimate the burning degree of the fire flame so that users could be informed with a proper warning alarm. The proposed method is tested with twelve diverse fire flame video clips at the process speed of thirty frames per second. The proposed method can achieve more than 97% detection rate on the average. In addition, the system can correctly recognize fire flames within one second on the initial combustion from the test video clips, which seems very promising. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|