基于YIQ颜色空间的火焰轮廓提取算法

A Method of Flame Contour Extraction Based on YIQ Color Space

为了能准确地分割图片中的火焰区域和背景区域并精准的提取火焰轮廓,本文提出基于YIQ颜色空间的火焰轮廓提取方法.该方法利用火焰的颜色和亮度特征,通过大量实验观察,在YIQ颜色空间,先获取Y、I分量的像素均值,再运用迭代法和灰度直方图法分别得到Y分量及I分量的最佳阈值,然后结合二者的均值对阈值进行取向处理,最后运用canny算子对得到的二值图进行轮廓提取.实验结果表明本文算法火焰分割精确度高,提取轮廓清晰,时间复杂度也得到很大改善.

In order to accurately divide the flame zone from the background and extract the flame contour of a flame picture. This article proposes a algorithm of flame contour extraction based on YIQ color space. This algorithm uses the color and brightness features of flame,by a large number of experimental observation,in the YIQ color space,to obtain firstly the mean value of Y and I component,then,the best threshold of Y and I by Histogram method and Iterative Method,and aligning treatment Y and I threshold combined with their mean value.At last,it uses the canny algorithm to extract contour of Binary image.Experimental results show that this method has high precision,more clear outline and less processing time in the field of flame contour extraction than other algorithm.