摘要
本文以蒙特卡罗模拟方法为基础,结合组织光学的光子传输模型,提出了一种新的图像分割算法,该算法将复杂的图像分割问题简化为大量简单的光子传输随机实验,通过分析传输规律来获取目标区域。在随后的实验中,结合细胞核提取这一问题建立了一个简单的光学传输模型,并依据此模型分别对人造图和实际图进行了分割。人造图的分割结果表明了该算法的可行性,说明了该算法的一些优点;而实际图的分割结果则反映了该算法的不足之处,文章针对其中存在的问题和算法待改进之处进行了分析。
A new segmentation method based on Photon propagation model is presented with the foundation of Monte Carlo Simulation. This algorithm simplifies the complexity of image segmentation by turning it into vast simple simulations of photon propagation, and the object is finally obtained by analyzing the law of propagation instead of the complex image data. In the later experiment, according to the problem of karyon segmentation, a photon propagation model is established, and segmentation works have been done on both artifact and real cell image. The result of artifact is good, and shows the feasibility of this new algorithm. Based on the real image' s results,existing problems have been analyzed.
出处
《激光生物学报》
CAS
CSCD
2009年第1期112-115,120,共5页
Acta Laser Biology Sinica
基金
厦门大学固体表面物理化学国家重点实验室开放课题
国家自然科学基金项目(60778046)
福建省科技项目(2008I0015
2008J0016)
关键词
图像分割
蒙特卡罗模拟
光子传输模型
细胞核提取
image segmentation
monte carlo simulation
photon propagation model
karyon extraction