基于Cr、Cb色彩空间的森林火焰图像分割方法

Forest flame image segmentation method based on Cr,Cb color space

为了解决对有很强日光干扰的火焰图像进行分割的问题,通过研究森林火焰在Y、Cr、Cb颜色空间上的分布情况,提出了一种基于Y、Cr、Cb空间的森林火焰彩色分割方法以及利用相似距离进一步提高处理效果的方法.实验表明,该处理方法对在有较强日光干扰下的森林火焰图像有良好的提取效果.该方法充分利用了火焰图像的分布特性以及颜色空间的色彩信息和不同颜色空间特征的互补性,为火焰图像的提取提供了一个新思路.

In order to solve the flame image segmentation problem with strong sunlight disturbance, the distribution of the forest flame in Y, Cr, Cb color space was investigated. A forest flame color segmentation method based on Y, Cr, Cb color space was proposed, and the similar distance method was adopted to enhance further the processing effect. The results indicate that the proposed processing method has better effect in segmenting the forest flame image with the stronger sunlight disturbance. The proposed method utilizes sufficiently the distribution feature of the flame image as well as the complementarity between the color information and characteristic in different color spaces, and provides a new view for flame image segmentation.