基于最佳缝合线和改进加权平均融合的图像拼接
Image Stitching Based on Optimal Seam Line and Improved Weighted Average Fusion
王金良 1闫素2
作者信息
- 1. 江西科技学院 南昌 330098
- 2. 潍坊工程职业学院 潍坊 262501
- 折叠
摘要
为了解决图像拼接中出现的拼接缝和伪影问题,文章提出一种基于动态规划思想的寻求最佳拼接缝算法,采用改进的加权平均融合算法解决拼接缝问题.在图像配准方面,通过SIFT算法对图像进行特征匹配,并利用RANSAC算法去除误匹配点.在图像融合方面,给定一对对齐的图像重叠区域,利用动态规划思想来搜索最佳缝合线;然后结合改进的加权平均融合算法实现无缝拼接.实验结果表明,本文提出的算法有效消除伪影和拼接缝,实现两幅图像重叠区域之间的光滑过渡,符合人类的视觉感知效果.
Abstract
In order to solve the problem of seams and artifacts in image stitching,this paper proposes an algorithm for finding the best seam based on the idea of dynamic programming,and uses an improved weighted average fusion algorithm to solve the seam problem.In the aspect of image registration,the SIFT algorithm is used to match the image features,and the RANSAC algorithm is used to remove the mismatched points.In terms of image fusion,given the overlapping regions of a pair of aligned images,dynamic programming is used to search for the best seam line;then combined with an improved weighted average fusion algorithm to achieve seamless stitching.The experimental results show that the algorithm proposed in this paper effectively eliminates artifacts and seams,and achieves a smooth transition between the overlapping regions of two images,which is in line with human visual perception.
关键词
动态规划/SIFT算法/最佳接缝/改进的加权平均融合Key words
Dynamic Programming/SIFT Algorithm/Best Seams/Improved Weighted Average Fusion引用本文复制引用
基金项目
江西省自然科学基金青年基金(20192BAB211005)
国家自然科学基金(61971234)
国家自然科学基金(61865012)
江西省研究生创新专项(YC2021-S821)
出版年
2024
NETL
NSTL
万方数据