基于交叉视觉皮质模型的图像快速分割新算法  

A novel image segmenting-promptly method based on intersecting cortical model

作　　者：朱孝政 

机构地区：[1]海军装备部,陕西西安710043

出　　处：《电子设计工程》2013年第8期47-50,共4页Electronic Design Engineering

摘　　要：为了使光学图像分割处理达到实时性标准,本文引入了具有生物学背景的交叉视觉皮质模型用于图像分割。在不降低分割效果的前提下,将交叉视觉皮质模型的阈值衰减函数调整为线性衰减函数,形成了线性衰减阈值-交叉视觉皮质分割新模型;引入了运算简便且抗噪性强的Chi-square散度,采用灰度级-邻域平均灰度级二维直方图将其推广至二维空间,设计了一种新的分割准则以确定模型的参数与循环迭代次数。Abstract： In order to meet with the real time standard of optical images ＇process, the intersecting cortical model in the biologic background is introduced for image segmentation. On one hand, the linear attenuation threshold function substitutes for exponential decay mechanism of the original intersecting cortical model on the occasion that the result should not be degenerated. Thus, the linear attenuation threshold-intersecting cortical model is established. On the other hand, the chisquare divergence with simple operation and strong noise immunity is extended to 2-D space by the gray level-average gray level 2-D histogram. Thus, a new segmentation rule is designed to get parameters and loop iterations.

关 键 词：图像分割 交叉视觉皮质模型 阈值衰减函数 灰度级-邻域平均灰度级二维直方图 Chi-square散度 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]