(X, Y, Z) world coordinates from (U, V) image coordinates, ROS


I’m developing a ROS node subscribing to the image data and the point cloud published by the ZED ROS node. With the image data, I perform object detection and get (U, V) coordinates describing the position of objects. From these (U, V) image coordinates, I would like to obtain the (X, Y, Z) world coordinates corresponding to that pixel.

I’m fairly new to Point Clouds, would appreciate help for this. Thank you!

Hi @Pedroacsilva
there is a 1-to-1 matching between the data in the RGB image matrix and the data in the point cloud matrix because they are registered, so if you take the i-th pixel in the RGB image you can get its 3D coordinates in the point cloud i-th pixel.