Overview of Multi-view Image Registration

Recognized as a foundational operation for advanced visual data analysis, image registration underpins critical innovations across fields such as computer vision and digital image processing. Our article introduces the definition of image registration and its application scenarios in different fields, then discusses kinds of image registration methods in detail, including Feature-Based Registration, Region-Based Registration, Deformable Registration, Intensity-Based Registration and Deep Learning-Based Registration. Through the analysis and summary of these methods, the existing problems of existing image registration methods are sorted out, and future research directions are prospected. This comprehensive discussion aims to promote continuous innovation in image registration technology to meet evolving application needs.