0

I have a depth map (500x290) and I want to obtain point cloud from it (x,y,z coordinates). The depth map data is produced by a perspective camera of 120 degrees horizontal field of view and 45 degrees of vertical field of view, therefore, the intrinsic parameter matrix is as follows:

         144.33    0     250
     K =   0     350.06  145
           0       0      1

the values in the depth map corresponding to the distance to the object (i.e. the norm of the vector (x, y, z)) and not only its Z component (depth). with all this information how can I calculate the x,y and z coordinates?

f3rhood
  • 5
  • 6
  • Possible duplicate of [Extracting 3D coordinates given 2D image points, depth map and camera calibration matrices](https://stackoverflow.com/questions/31265245/extracting-3d-coordinates-given-2d-image-points-depth-map-and-camera-calibratio) – serkan.tuerker Jan 16 '19 at 21:58
  • nope. it is not the same thing. the sensor is not Kinect. – f3rhood Jan 18 '19 at 13:59
  • that doesn't matter. check out the formula how to compute the X,Y,Z values. it is the same concept. – serkan.tuerker Jan 18 '19 at 19:28
  • I have tried and it doesn't work. this is my data: https://drive.google.com/open?id=1NlH7fxkSOZjcVSrtg2CB8gcUGFDONG21 and this is the depth image: https://drive.google.com/open?id=14zJoJku7Y7-gs7iaGtReLq8scy-uSMeE let me know if you could correctly construct the scene without distortion (with the abovementioned information of the sensor). good luck. – f3rhood Jan 19 '19 at 21:02

0 Answers0