Abstract: Image information contains a lot of uncertainty which may cause the unsatisfactory image segmentation effect. In order to solve this problem, this paper proposes a algorithm of color image segmentation based on granular computing and cloud model. It granulates color images in HSV color space, and cloud-granule concept is generated in multi-granularity and multi-level, finally the color image is segmented. Experimental results show that the proposed algorithm has better performance compares with PCNN and K-means algorithm.

Key words: granular computing, cloud model, color image segmentation, cloud transformation, multi-granularity, HSV space

摘要： 图像信息中存在的不确定性问题会影响图像的分割效果。为此，提出一种基于粒计算和云模型的彩色图像分割算法。研究多粒度认知模型，在HSV颜色空间中利用云模型构建彩色图像的信息粒，进行多粒度、多层次的云粒合成，实现彩色图像分割。实验结果表明，与PCNN算法和K均值算法相比，该算法的分割效果较好。

关键词: 粒计算, 云模型, 彩色图像分割, 云变换, 多粒度, HSV空间

CLC Number: 

• TP391