基于椭圆拟合和Hausdorff距离的图像匹配算法  

Image Matching Algorithm Based on the Ellipse Fitting and Hausdorff Distance

作　　者：董方[1] 胥宏[2] 胡亮[1] 方黎勇[1] 李柏林[1] 

机构地区：[1]西南交通大学机械工程学院,成都610031 [2]成都电子机械高等专科学校成人教育学院,成都610031

出　　处：《CT理论与应用研究（中英文）》2009年第1期41-46,共6页Computerized Tomography Theory and Applications

基　　金：四川省科技攻关资助项目(04GG010-004-1)

摘　　要：图像匹配是逆向工程CAD建模领域的重要内容之一。针对CT图像包含多分叉轮廓的匹配问题,利用椭圆拟合来定量表达平面轮廓散乱点的整体形位信息,建立初步的轮廓森林;提出采用快速Hausdroff距离法来处理分枝问题从而形成完整的轮廓森林。实验结果表明,该算法对具有分叉现象的图像匹配具有良好的准确度和实时处理能力,对于岩芯CT、工业CT、医学CT等图像的三维重构都能有较好的适用性。Image matching is an important part of the field which is about reverse engineering CAD modeling,For the matching problem of CT images which contain bifurcation outline, ellipse-fitting was used to quantitative describe the completed shape and locating information for the series of discrete contour points ,thus a elementary contour forest can be established. Speedy HD distance method was used to deal with the issue of branches to form a complete contour forest. Experimental results show that the algorithm to imatch images with good accuracy and real-time processing capability. It is useful to match CT images of rock heart, industry and medical.

关 键 词：轮廓匹配 轮廓森林 椭圆拟合 分枝 HAUSDORFF距离 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]