Fire Image Segmentation Algorithms

In the large space environment, the traditional fire detection technology is difficult to work properly because the large space environment is complex and changeable and the impact of interference is serious. The image-based fire detection technology is a kind of new detection technology, which is composed of fire image segmentation, fire feature extraction, fire recognition and automatic fire extinguishing technology. The fire image segmentation is the premise of the fire feature extraction and recognition, which will decide the accuracy of the fire recognition. Therefore the research of the fire image segmentation is very important.A modified FCM algorithm for fire image segmentation is proposed in this paper based on the study of fuzzy clustering algorithm and characteristics of fire images. The new algorithm is based on HSI color space, initial clustering centers of H component and I component are selected according to the data distribution. Then, clustering for H component and I component is implemented in histogram feature space, the membership function is improved with the pixel's spatial feature. Finally, the algorithm is performed in the two-dimensional features space which is constructed with the image pixel membership for H component and I component. The results show that the new algorithm make full use of fire color information to solve the problem of the inaccurate segmentation which is caused of the difficulty to determine the experience threshold and the losting of the color information. The proposed method can extract flame region accurately and has real-time character, which is very important for the fire detection.