Helmet Detection Using YOLO V3 And Single Shot Detector

Wearing a Safety helmet is vital to the well-being of labourers at building destinations and production lines. Instructions to caution or recognize labourers regardless of whether the protective helmet is worn, is regularly a troublesome point for undertakings to screen. This project presents a mechanized system to recognize motor bikers without a helmet and labourers without wearing the helmet. On the helmet identification conundrum, a You Only Look Once (YOLO) model is used. This model will only consider the bounding box territories of construction labourers and motorcyclists using a single structure. Further applying the Single-shot detector (SSD) model to upgrade the prepared pictures likewise perform better in picture limitation because of successful component combination, which is more under the prerequisite of acknowledgement of helmet. The basis of this model is highly critical in terms of enhancing helmet recognition and ensuring secure estimation.