CAJ | 학술논문

模糊C均值（FCM）被广泛应用于彩色图像分割中，但传统的模糊C均值由于没有考虑空间信息，因此对噪声特别敏感。针对此问题，提出了一种在HIS颜色空间结合像素邻域空间信息的模糊聚类新方法。实验结果表明，此方法对高噪声图像有较好的处理结果。
모호C균치（FCM）피엄범응용우채색도상분할중，단전통적모호C균치유우몰유고필공간신식，인차대조성특별민감。침대차문제，제출료일충재HIS안색공간결합상소린역공간신식적모호취류신방법。실험결과표명，차방법대고조성도상유교호적처리결과。
Fuzzy C-Mean（FCM） algorithm has been wildely used in the field of color image segmentation. But the convention- al FCM is sensitive to the noise because the spatial information is not being considered. To solve this problem, a new fuzzy clustering algorithm in HIS color space is proposed based on pixel＇s spatial neighborhood and pixel value. Experimental results prove that it has a good result to high noise image.