摘要
在组织工程培养细胞图像检测中,需要得到较完整的清晰轮廓线以获取生长信息。为了适应这种需求,提出了一种基于Snake算法自动边缘检测算法。应用基于小波的尺度独立边缘检测方法,可以得到简洁的边缘;采用形态学和样条拟合方法得到的单像素的边界作为Snake算法的初始轮廓,通过迭代和结合图像固定点的外部约束力使轮廓收敛到最终的边界。实验结果证明了该算法的可行性和有效性。
An edge detection algorithm based on snake active contour for getting clear contour line of cell in tissue engineering is proposed. First, rough edge by scale_independ algorithm of wavelet transform is got, then morphological and sample methods are adopted to get single pixel closed boundary as initial contour of snake algorithm, finally contour converge is made to the actual boundary through iterant arithmetic and exterior sanction that combines image fix point. The validity and feasibility of this algorithmis proved by the experiment.
出处
《计算机工程与设计》
CSCD
北大核心
2006年第21期4153-4156,共4页
Computer Engineering and Design
关键词
组织细胞
小波变换
图像分割
活动轮廓
蛇型模型
tissue cell
wavelet transform
image segmentation
active contour
snake model
