基于YCbCr空间对大壁虎脑图谱的分割  

Gekko Gecko′s Brain Atlas Segmentation Based on YCbCr Space

作　　者：孟龙晖[1,2] 王文波[1] 戴振东[1] 

机构地区：[1]南京航空航天大学仿生结构与材料防护研究所,江苏南京210016 [2]南京航空航天大学机电学院,江苏南京210016

出　　处：《现代电子技术》2011年第4期84-86,90,共4页Modern Electronics Technique

基　　金：国家自然科学基金重点项目(60535020);国家自然科学基金面上项目(30700068)

摘　　要：确定大壁虎的各种运动行为与脑内控制其运动的具体的皮层组织对应关系是壁虎机器人的制作关键。通过对其脑图谱的制备,可以观察到脑内皮层组织的丰富信息。针对HE染色的大壁虎脑图谱自身的独特性,经过推敲,提出一种基于YCbCr空间的对图谱进行分割的方法。该方法将图谱由RGB空间转换为YCbCr空间,将亮度分量独立出来,不仅解决了拍摄过程中由于光线改变所带来的问题,而且阈值分明,分割彻底,鲁棒性好。其结果表明基于YCbCr空间的分割方法是分割效果非常好的方法。The key to make the bio-robot is to identify the relationship between the Gekko Geeko＇s behavior and the corresponding cortex organization of its brain. By the way of making its brain atlas, a wealth of information of cortex organization can be observed. As the feature of Gekko gecko＇s brain atlas dyed by HE method, after scrutiny, a method of segmentation based on YCbCr space is proposed. This method converts the atlas from RGB space to YCbCr space and makes the luminance component independent. This method not only solved the problem caused by the changing light, but also make the threshold clear and the segmentation complete. It has a good immunity to the noise. The results show that the method based on YCbCr space is indeed a good method to make the segmentation well. The implementation process was achieved by the programs based on Matlab.

关 键 词：大壁虎脑图谱 YCBCR空间 大脑图谱分割 阈值 

分 类 号：TN919-34[电子电信—通信与信息系统]