Estimate Depth from a Single Image

Use a neural network that was trained to predict the relative depth map from a single image using a technique based on sparse ordinal annotations:

In[1]:=

net=NetModel["Single-Image Depth Perception Net Trained on NYU Depth V2 and Depth in the Wild Data"]

Out[1]=

Get a 2D image:

In[2]:=

image=Import["https://cdn.pixabay.com/photo/2017/02/14/03/03/ama-dablam-2064522_1280.jpg",ImageSize340]

Out[2]=

Apply the net to the image to get the 2D depth map. Brighter points correspond to further regions:

In[3]:=

depthMap=net[image];Image[depthMap]//ImageAdjust

Out[3]=

Use

to plot a 3D representation of the depth map:

In[4]:=

ListPlot3D-Reverse@Normal@depthMap,



Out[4]=

Combine the original image with the depth plot using the

directive:

In[5]:=

ListPlot3D-ArrayResample[Reverse@Normal@depthMap,Reverse[ImageDimensions[image]]],PlotStyleTexture[image],



Out[5]=

Wolfram Language Example Repository