In view of the selection of structure elements problem in morphological filter,this paper presents morphological operators with non-fixed shape and non-fixed size kernels,or amoebas,instead of fixed-shape,fixed size structuring elements generally used.These structure kernels take into account the image contour variations to adapt their shape and size according to the gradient criteria built firstly.Then,it performs the median or mean filtering at each pixel with these structure kernels.Experiments results show that this method can simultaneously suppress the clutter and enhance edges to overcome the transitional filter′s drawbacks,simultaneously comparing the anisotropic filter has better efficiency.
参考文献
相似文献
引证文献
引用本文
王艳华,刘伟宁,陈爱华.基于数学形态学变形虫的自适应图像滤波[J].激光与红外,2008,38(12):1258~1261 WANG Yan-hua, LIU Wei-ning, CHEN Ai-hua. Adaptive Image Filter Algorithm Based on the Mathematical Morphology Amoebas[J]. LASER & INFRARED,2008,38(12):1258~1261