CPC G06T 15/60 (2013.01) [G06N 3/045 (2023.01); G06T 7/194 (2017.01); G06T 7/70 (2017.01); G06T 11/00 (2013.01); G06T 2200/24 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2215/12 (2013.01)] | 20 Claims |
1. A method comprising:
extracting, from a digital image portraying a digital object, an object mask corresponding to the digital object;
generating, utilizing a height prediction neural network, a height map from the object mask corresponding to the digital object, the height map comprising pixel heights for pixels of the digital object;
determining vertical projections of the pixels of the digital object onto a surface portrayed in the digital image using the pixel heights for the pixels in the height map;
determining shadow projections for the pixels of the digital object using the pixel heights from the height map and a light source for the digital image by determining, as a location of a shadow pixel corresponding to a pixel of the digital object, an intersection between a first line that includes the light source and the pixel and a second line that includes a vertical projection of the light source onto a horizon of the digital image and a vertical projection of the pixel onto the surface determined using the height map; and
modifying the digital image to include an object shadow for the digital object using the shadow projections for the pixels of the digital object.
|