CPC G06T 17/205 (2013.01) [G06T 7/60 (2013.01); G06V 10/26 (2022.01); G06V 10/764 (2022.01); G06T 2207/10016 (2013.01); G06T 2207/10024 (2013.01); G06T 2207/10028 (2013.01); G06T 2210/56 (2013.01)] | 20 Claims |
1. A method, comprising:
receiving, from a capturing device, frames of color data and frames of depth data;
generating, from the frames of depth data, a 3D mesh representation of a 3D space;
classifying pixels of the frames of color data as representing an object in the 3D space based on a semantic segmentation process, to generate semantic segmentation images, wherein the semantic segmentation images define one or more boundaries of the object;
projecting the semantic segmentation images to the 3D mesh representation; and
updating the 3D mesh representation of the 3D space based on the one or more boundaries in the projected semantic segmentation images, wherein faces of the 3D mesh representation comprise polygons and the updating comprises:
fitting a line to the one or more boundaries;
determining a point at which the line intersects at least one of the polygons; and
splitting the at least one of the polygons into multiple new polygons having a vertex at the point.
|