广西大学学报（自然科学版）

彩色图像分割一直是彩色图像处理与分析中最为困难的不可缺少的步骤,针对图像分割质量直接在很大程度上影响了图像后期分析的效果,提出了一种基于HSI和LAB颜色空间的彩色图像分割方法。该方法在HSI颜色空间用最优阈值方法进行阈值分割,在LAB颜色空间采用基于K均值聚类图像分割,然后将两次分割结果进行区域合并,最后进行加窗滤波消除噪声。对林区活立木真彩色图像进行分割的实验结果表明,该方法能够精准地将活立木从背景中提取出来。

The color image segmentation is the most difficult and undependable phase in color image processing,which might directly determine the results of final analysis to a great extend.A method of color image segmentation based on HSI and LAB color space is proposed in this paper.Firstly,the method uses the means of optimum threshold segmentation in HSI color space and K-means clutering segmentation in LAB color space.And then,the method combines two of segmentations together as a whole and filters the noise by window function.The segmentation experiment on a real color image of forest shows that the proposed method can Abstract the trees precisely from the background.