Title :
Classification of flame and fire images using feed forward neural network
Author :
John, Olga ; Prince, Shajin
Author_Institution :
Dept. of Electron. & Commun., Karunya Univ., Coimbatore, India
Abstract :
Flame edge determination and classification is an active research area of wide application. Edge detection techniques for flame images can be used in applications such as flame image monitoring, early detection of fire, evaluation of fire, and the measurement of flame and fire parameters. This paper mainly gives importance on edge detection and classification of flames. The boundaries of flames are detected perfectly by using auto adaptive edge detection algorithm and classification of flame levels are done by using resilient back propagation algorithm in feed forward neural network.
Keywords :
backpropagation; edge detection; feedforward neural nets; fires; flames; image classification; auto adaptive edge detection algorithm; edge detection techniques; feed forward neural network; fire early detection; fire evaluation; fire images; fire parameters; flame edge classification; flame edge determination; flame image monitoring; flame levels classification; flame measurement; flames boundaries; resilient back propagation algorithm; Classification algorithms; Convolution; Feedforward neural networks; Fires; Image edge detection; Edge detection; edge image analysis; feature extraction; flame classification; shape measurement;
Conference_Titel :
Electronics and Communication Systems (ICECS), 2014 International Conference on
Conference_Location :
Coimbatore
Print_ISBN :
978-1-4799-2321-2
DOI :
10.1109/ECS.2014.6892629
