火焰的红色饱和度-亮度模型

摘要

By studying images of different kinds of flames, the spatial distribution of their luminance and the change regularity of red-saturation with luminance between different flame pixels are found, and the Red-saturation Luminance (RL) model of flames are presented. This model describes the change regularity of pixel parameters among different pixels. In the application of fire detection, suspected flames can be extracted by flame segmentation, and then judged by RL model whether the fire accidents are happened or not.%通过研究不同的火焰图像,发现了火焰图像亮度的空间分布规律及不同像素之间红色饱和度随亮度的变化规律,提出了火焰图像的红色饱和度亮度(RL)模型.该模型描述了火焰图像的像素参数在不同像素间的变化规律.在在火灾探测应用中,可通过图像分割将疑似火焰图像从摄像头拍摄的视频中提取出,通过RL模型判定是否发生了火灾.