RA-MMIR : Multi-modal image registration by Robust Adaptive Variation Attention Gauge Field

Junhui Qiu, Hao Li, Hualong Cao, Xiangshuai Zhai, Xuedong Liu, Meng Sang, Kailong Yu, Yunpin Sun, Yang Yang, Pan Tan

Abstract: Highlights•We propose a Weak Boundary Constraints-based Flexible Matching strategy.•We propose a RAMM Attention to perform feature extraction and description.•We propose a Unified-VEM to minimize computational complexity.•We propose a SP-SSIM making transformation more consistent with real-world scene.