Abstract: The conventional security protective cap wearing acknowledgment is just founded on the shading, shape, surface and different attributes of the picture, which is extraordinarily impacted by the outer climate, and has the issue of temperamental acknowledgment precision. Considering the above issues, this paper concentrates on the acknowledgment strategy for power development labourer’s security head protector dependent on man-made brainpower innovation. Subsequent to pre-processing the development checking picture, for example, turning Gray and denoising, the development work force in the identification picture are found, that is, based on recognizing the development staff region, the head position of the development faculty is found, lastly the wellbeing protective cap wearing acknowledgment is acknowledged by utilizing YOLO calculation. The re-enactment results show that the normal acknowledgment exactness is 95.2%, the acknowledgment impact is steady and has great heartiness.
Keywords: Helmet detection, Construction sites, Deep learning, Accidents
|
DOI:
10.17148/IJARCCE.2022.11128
