A Dual-Modality Pedestrian Detection Method Based on Multi-Scale Feature Fusion

In urban safety, autonomous transportation, and intelligent surveillance, advanced pedestrian detection technology is crucial. Current methods based on visible light imaging face limitations in low-light or adverse weather conditions, leading to reduced detection accuracy. To address these challenges, this paper introduces a novel dual-modality pedestrian detection method using the YOLOv7 model, enhanced with Modality Alignment (MA) and Differential Modality Fusion (DMF) modules. These modules efficiently utilize dual-modality data, combining visible light and infrared imaging, to improve detection in various environmental conditions. An integrated Mix module further enhances feature extraction and fusion. The proposed approach demonstrates a marked improvement in pedestrian detection accuracy across diverse scenarios, offering a more robust and adaptive solution for complex urban and rural environments. This advancement signifies a significant leap in pedestrian detection, leveraging multi-scale feature fusion and innovative algorithmic strategies.