Research on 24-Hour Dense Crowd Counting and Object Detection System Based on Multimodal Image Optimization Feature Fusion

Table 5

Comparison of the different state-of-the-art target detection methods in day and night.

Methods	Type	Dataset	Day (AP)	Night (AP)	Day + night (AP)
DetNet [71]	RGBD	ShanghaiTechRGBD	0.383	—	—
RDNet [71]	RGBD	ShanghaiTechRGBD	0.610	—	—
CGD [71]	RGBD	ShanghaiTechRGBD	0.727	—	—
YOLO4-RGB [72]	RGBT	KAIST	0.684	0.298	0.465
YOLO4-T [72]	RGBT	KAIST	0.641	0.617	0.625
YOLO4-RGBT [72]	RGBT	KAIST	0.648	0.609	0.618
Ours	RGBD	ShanghaiTechRGBD	0.825	0.806	0.816
Ours	RGBT	KAIST	0.726	0.715	0.721