一种基于块结构特征的病理切片图像拼接算法  

Stitching Image Algorithm for Pathologic Slices Based on Block Structure Characteristics

作　　者：曹红苹[1] 王国钧[1] 刘建华[2] 

机构地区：[1]湖州师范学院信息工程学院,浙江湖州313000 [2]桂林航天工业高等专科学校,广西桂林541004

出　　处：《桂林工学院学报》2005年第3期364-366,共3页Journal of Guilin University of Technology

基　　金：国家自然科学基金资助项目(60275023)

摘　　要：提出了一种抽取简单而结构性能较强的块结构特征作为模板的新的病理切片图像拼接算法,由于引入了图片的相关性和它的最优化处理,并利用最大包的分布进行图像配准,所以算法的速度快、稳定性高.图像的重叠区可以只取100列宽度,实验取得了较好的配准效果.This introduction is about a new pathological cut-section image stitching algorithm which uses a simple and powerful structural block as a template. The algorithm refers to the stitching of two adjacent pathological cut-section images in the overlapping area, a further development of Mr. Zhong Li＇s algorithm. With the correlation and the maximum packet distribution for image mosaic, the algorithm is faster and more stable to stitch images with 100 columns only in the overlapping area. A more practical effect can be achieved.

关 键 词：结构特征 相关性 图像拼接 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]