How to combine 2d and 3d information?

asked 2017-11-19 21:05:27 -0500

Ted.Xie gravatar image

Hey all :

I want to use line laser to scan my object and use camera to find object feature. First, I will use camera to find my object than get the object contour, but now only have x,y information, so i want to join line laser to scan the same object to get 3d information, what should i do to find the 2d object contour in 3d depth. I want to combine two different information.

Thanks in advance

edit retag flag offensive close merge delete


Like you said if you have a laser device you can simply bounce it off the object to measure distance. You can then augment that distance (Z) value with X,Y value that you recover from (x,y) of image using your camera calibration matrix.

ZainMehdi gravatar image ZainMehdi  ( 2017-11-21 06:58:09 -0500 )edit

Thank you! How should i know the image(x,y) and point cloud(x,y) is same,because it is different in formation?

Ted.Xie gravatar image Ted.Xie  ( 2017-11-22 01:53:26 -0500 )edit

If you use kinect then you dont need this reprojection. Kinect itself will give you depth information along with spatial coordinates.

ZainMehdi gravatar image ZainMehdi  ( 2017-11-22 01:56:20 -0500 )edit

But kinect accuracy is not good,so i am use line laser and camera.

Ted.Xie gravatar image Ted.Xie  ( 2017-11-23 00:42:37 -0500 )edit

lets say you measure some feature at some random position in your image, you can just point your laser at that location to get the depth and you already have x and y using your calibrated camera. It might be that you cant accurately point your laser at particular feature location but you can *

ZainMehdi gravatar image ZainMehdi  ( 2017-11-26 11:58:18 -0500 )edit

generalize it. Like your features on on some object like a table so you can point your laser on that table and get the depth data.

ZainMehdi gravatar image ZainMehdi  ( 2017-11-26 11:59:09 -0500 )edit