Speed Matters, a robust infrared and visible image matching method at real-time speed

Rong Chang,Chuanxu Yang,Hang Zhang,Housheng Xie,Chengjiang Zhou,Anning Pan,Yang Yang

DOI: https://doi.org/10.1007/s11554-023-01395-4

IF: 2.293

2024-01-04

Journal of Real-Time Image Processing

Abstract:Image matching is a crucial step in executing many complex visual tasks, which involve identifying the same or similar visual patterns across various images. Matching images between infrared and visible becomes quite challenging due to the significant non-linear intensity differences. In this paper, we propose using a lightweight network for feature matching of infrared and visible images, combining global and local feature information, and reducing computational costs, enabling real-time inference on most desktop-level GPUs. To fully leverage the powerful matching capabilities of existing state-of-the-art models, we introduce knowledge distillation to obtain more robust features. Moreover, to address the issue of insufficient datasets for network training in existing methods, we propose using image style transfer techniques to synthesize paired datasets of infrared and visible. Experimental results show that our method achieves results comparable to the most advanced methods in infrared and visible image matching. Furthermore, our method has a significant advantage in inference speed, which is beneficial for tasks that require real-time completion.

computer science, artificial intelligence,engineering, electrical & electronic,imaging science & photographic technology