摘要
针对现有SIFT算法在图像拼接技术应用中存在的不足,提出了改进的分块H-SIFT算法.利用SIFT算法的可扩展性,与Harris算法联合使用,并对图像进行分块配准,在提高配准精度的同时,提高配准效率.在加权平滑的基础上,利用图像特征点处的梯度、散度值寻找最佳拼接线,实现图像的无缝融合.
In view of the shortcomings of existing SIFT algorithm in the application of image mosaic technology,an improved block H-SIFT algorithm was proposed in this paper.The SIFT algorithm was combined with Harris algorithm,and the images was blocked before image matchting,which improved the efficiency and accuracy of image matchting.On the basis of weighted smoothing algorithm,the gradient and divergence value of the image feature points were used to find the best splicing line and realize the seamless mosaic of the images.
作者
李旋宇
陈荣保
LI Xuanyu;CHEN Rongbao(School of Mechanical and Electrical Engineering,Chizhou University,Chizhou 247000,China;School of Electrical and Automation Engineering,Hefei University of Technology,Hefei 230009,China)
出处
《河南科技学院学报（自然科学版）》
2018年第2期48-51,59,共5页
Journal of Henan Institute of Science and Technology(Natural Science Edition)
基金
池州学院自然科学重点项目(2016ZRZ005)
