摘要
文章提出一种基于SIFT改进的图像拼接算法。通过小波变换对图像进行预处理,提取图像的SIFT特征点生成描述符并进行匹配,使用RANSAC算法去除图像匹配中错误的匹配点对并进行图像融合完成图像拼接。实验结果表明,利用文章提出的算法对图像进行提取特征平均耗时短,拼接效果好,没有明显裂痕。
In this paper,an improved image Mosaic algorithm based on SIFT is proposed.The image is preprocessed by wavelet transform.Then,the SIFT feature points of the image are extracted to generate descriptors and matched.Finally,the RANSAC algorithm is used to remove the wrong matching point pairs in the image matching,and the image fusion is completed.The experimental results show that,the algorithm proposed in this paper takes less time to extract the features of the image,and the mosaic effect is good,without obvious cracks.
作者
张文静
刘茵
侯英竹
李明亮
ZHANG Wenjing;LIU Yin;HOU Yingzhu;LI Mingliang(Department of Information Technology,Beijing Vocational College of Agriculture,Beijing 102442,China;College of Information Engineering,Hebei GEO University,Shijiazhuang 050031,China;Intelligent Sensor IOT Engineering Research Center of Hebei Province,Shijiazhuang 050031,China)
基金
北京农业职业学院科技创新项目基于VR技术的360全景展示设计及实现(No.XY-YF-20-13)
北京市职业院校教师素质提升计划专业创新团队培养项目(设施农业与装备(智慧农业)专业创新团队)。
关键词
小波变换
SIFT算法
RANSAC算法
图像拼接
wavelet transform
scale-invariant feature transform
random sample consensus
image mosaic
