基于在线生物学文献的MRI图像获取  

MRI image extraction from online biological literature

作　　者：赵鹏飞[1] 钱沄涛[1] 郑文斌[1] 

机构地区：[1]浙江大学计算机学院,浙江杭州310027

出　　处：《中国体视学与图像分析》2009年第2期216-221,共6页Chinese Journal of Stereology and Image Analysis

基　　金：国家自然科学基金资助(60872071)

摘　　要：近年来,利用在线文献构建生物数据库引起越来越多的关注,它包括论文中生物数据的自动收集、组织和分析。在线文献中图像形式表示的数据具有特别的重要意义。而理解图表内容的第一步是分解图表为嵌图,并进一步分解嵌图为感兴趣的对象。作为从在线文献中构建核磁共振图像(MR I)数据库的必要部分,本文研究了嵌图提取以及MR I图像分割方法。提出了嵌图的递归分割,基于MR I图像检测的高斯混合模型,和基于区域增长的形态学算子方法。实验表明MR I或其它感兴趣的图像可以从文献的图表中自动精确地获取。它为我们构建能解释在线文献中MR I图像的知识系统这一长期目标提供了基础支持。Recently, building biological databases from online literature has attracted more attentions, which includes automating the collection, organization and analysis of biological data in the articles. Data in the form of images in online literature present special challenges for such efforts. The first steps in un- derstanding the contents of a figure are decomposing it into panels and even decomposing a panel into the interested objects. In this paper, as a necessary part of constructing the magnetic resonance imaging （MRI） database from online literature, the panel extraction and MRI image separation are studied. The recursive panel splitting, Gaussian mixture based MRI image detection, and morphologic operator based region growing methods are proposed. The experiments show that MRI or other type of interested images can be obtained precisely and automatically from the figures in the literature. It provides a foundation for our long-term goal of building a knowledge base system that can interpret MRI images in online articles.

关 键 词：MRI 在线文献 图像分割 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]