Local-adaptive and outlier-tolerant image alignment using RBF approximation
作者：
Highlights：
• Natural images usually cannot be accurately aligned using traditional approaches.
• Up-to-date approaches suffer from unreliable matches and high computational cost.
• Radial basis function approximation is used to efficiently align natural images.
• Incorrect matches are removed by analyzing parameters of the approximation.
• Overfitting issues are addressed by smoothly weighting the image deformations.
摘要
•Natural images usually cannot be accurately aligned using traditional approaches.•Up-to-date approaches suffer from unreliable matches and high computational cost.•Radial basis function approximation is used to efficiently align natural images.•Incorrect matches are removed by analyzing parameters of the approximation.•Overfitting issues are addressed by smoothly weighting the image deformations.
论文关键词：image alignment,radial basis function,scattered data approximation,outlier removal,computer vision
论文评审过程：Received 7 March 2019, Revised 12 January 2020, Accepted 28 January 2020, Available online 5 February 2020, Version of Record 18 February 2020.
论文官网地址：https://doi.org/10.1016/j.imavis.2020.103890
