重庆理工大学学报（自然科学版）2017，Vol.31Issue(8)：140-144,5.DOI:10.3969/j.issn.1674-8425(z).2017.08.023
一种基于邻域对比的拼接缝消除方法
Splicing Seam Elimination Method Based on Neighborhood Comparison
摘要
Abstract
For finding splicing seam with low accuracy and it can't eliminate splicing seam effectively in the image splicing, a splicing seam elimination method based on neighborhood comparison is proposed.First of all, use a 3×3 sliding template to find a pixel point whose gray standard deviation value which is the largest in each row of number of rows continuously from the image overlap areas line by line, and to exclude the pixel points with large deviation until after two points in iteration, and to calculate the mean value of the rest of points as the starting splice point of the splicing seam, and to set the threshold value to limit the range of candidate splicing point in turn on the basis of their column position.Then, it selects the points with largest gray standard deviation of the candidate points in the 3×3 neighborhood as the splice point of the row.Finally, it connects the optimal splicing point of each row as the optimal splicing seam.Experimental results show that the method can make the gray difference smaller on both sides of the splicing seam in order to eliminate the splicing seam relatively effectively, and to make the image overlap region smooth and natural, and to realize seamless splicing better.关键词
拼接缝消除/灰度标准差/最佳拼接缝/图像拼接Key words
splicing seam elimination/gray standard deviation/optimal splicing seam/image splicing分类
信息技术与安全科学引用本文复制引用
罗永涛,张红民,王艳,张见双,陈柏元..一种基于邻域对比的拼接缝消除方法[J].重庆理工大学学报（自然科学版）,2017,31(8):140-144,5.基金项目
重庆市基础与前沿研究计划项目(cstc2015jcyjA40051) （cstc2015jcyjA40051）
