"深度学习多模图像配准方法研究综述及性能比较"

age scene and facilitate subsequent visual tasks. With the development of computer power and the rise of deep learning, multimodal image registration methods continue to evolve, improving the registration performance. This paper provides a comprehensive discussion and analysis of different multimodal image registration methods and their development process. First, the basic concepts of image registration are introduced, and the fundamental ideas of various methods are comprehensively expounded, discussing the characteristics of various methods. Secondly, the limitations of various algorithms are summarized, and further directions for improvement are provided. Finally, specific definitions of various evaluation indicators are given, and the performance of each registration method is compared from multiple perspectives such as qualitative evaluation, quantitative evaluation, and operational efficiency. Keywords: Image registration; Multimodal image; Deep learning; Overview of multimodal image registration research.