Fusion of Multi-Intensity Image for Deep Learning-Based Human and Face Detection

For ordinary IR-illuminators in nighttime surveillance system, insufficient illumination may cause misdetection for faraway object while excessive leads to over-exposure of nearby object. To overcome these two problems, we use the MI3 image dataset, which is established by multi-intensity IR-illumination (MIIR), as our benchmark dataset modern detection methods. We first provide complete annotations its current ground-truth incomplete. Then, illuminated IR videos evaluate several widely used detectors, i.e., SSD, YOLO, Faster R-CNN, and Mask analyzing effective range different intensities. By including a tracking scheme, well developing new fusion method intensities improve performance, proposed approach serve face wide distances. The (Dataset available: https://ieee-dataport.org/documents/mi3 ) with more source codes (Codes are xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/thesuperorange/deepMI3 available online.