Abstract:

Based on human visual attention mechanism, this paper puts forward a new method for fire detection and identification by fusing conspicuousness attention mechanism. Namely seeing is believing, through simulation of the human visual system,fusing significance attention mechanism to extract important information from the images, and to eliminate redundant data,identify the suspected fire area. Then to use multiple types Support Vector Machine for training and learning, the experimental results show that the proposed detection method of fusing conspicuousness attention mechanism combined with the SVMs of multiple function kernels has high accuracy, good stability and the excellent overall performance and so on characteristics.

Key words: conspicuousness attention mechanism, suspected region;support vector machine, video images, fire detection