A Vision-Based Collision Monitoring System for Proximity of Construction Workers to Trucks Enhanced by Posture-Dependent Perception and Truck Bodies’ Occupied Space

In the study, an automated visualization of proximity between workers and equipment is developed to manage workers’ safety at construction sites using convolutional-neural-network-based image processing a closed-circuit television video. The images are analyzed automatically transform hazard index visualized in form plane map. graphical representation personalized map proposed termed as ellipse study. depending on posture area occupied by hazardous objects (trucks) enable represent precise proximity. Collision monitoring with computer vision techniques artificial-intelligence-based object detection, space calculation, pose estimation, homography.