一种改进的边缘停止函数图像分割方法

摘要

在基于边缘的几何活动轮廓模型中,边缘停止函数的选择会影响图像的分割效果,根据边缘停止函数的性质,提出一个新的边缘停止函数,代替GAC模型和DELSE模型中的边缘停止函数.实验结果表明,在图像背景灰度均匀和灰度不均的图像分割中,文中提出新的边缘停止函数在迭代次数和分割效果方面,都优于原来的边缘停止函数,能够实现准确分割.%In edge-based geometric active contour model,the choice of edge stop function affects the image segmentation effect.According to the nature of the edge stopping function,a new edge stopping function is proposed to replace the edge stopping function in GAC model and DELSE model.The experimental results show the image background segmentationwith uniform gray Scale and uneven gray scale a new edge stopping function is proposed,which is better than original edge stopping function in terms of iterations number and segmentation effect,and new edge stopping function can achieve accurate segmentation.