This project is based on the OpenCV perspective transformation. The goal is to find the real-world distance between objects (e.g., workers, equipment) or locating objects in specific zones using pixel coordinates and camera calibration parameters. Capture.png file shows an example of the proposed method. The average error measured for this project is around 1.5 m. To get the world coordinates run
Main.py
wehere, (100, 200) are the pixel coordinates of the equioment
equipment_world_coordinate = eq.transfer_2d_to_3d(100, 200)
The image bellow shows perspective locating of equipments on the construction site.