Converts an image into 3d mesh using depth map of the image.

• Opencv - Required for reading image
• Numpy - Required for storing image & for matrix operations

$ python depthTo3d.py <input-image> <input-depth> <output-ply-file> <fx> <fy> <cx> <cy>

$ python depthTo3d.py sofa.jpg sofa_depth.png s.ply 1148.93617021 1150.38461538 750 500

• Use texture mapping
• Smooth generated meshes
• Estimate position of the lights in the scene
• Estimate material properties
• Extract depth using single image

  from reconstruct import reconstruct3d
  from ply import *

  mesh = reconstruct3d(image,depth_map,camera_parameters, step = 1, mesh = True, transformation = None)

  mesh.save('out.ply', color = 'texture')