US 11,790,605 B2
3D reconstruction with smooth maps
Serban Alexandru State, Velizy Villacoublay (FR); Eloi Mehr, Velizy Villacoublay (FR); and Yoan Souty, Velizy Villacoublay (FR)
Assigned to DASSAULT SYSTEMES, Velizy Villacoublay (FR)
Filed by DASSAULT SYSTEMES, Velizy Villacoublay (FR)
Filed on Dec. 31, 2020, as Appl. No. 17/139,121.
Claims priority of application No. 19306796 (EP), filed on Dec. 31, 2019.
Prior Publication US 2021/0201571 A1, Jul. 1, 2021
Int. Cl. G06T 17/20 (2006.01); G06F 30/23 (2020.01); G06T 5/00 (2006.01); G06F 30/17 (2020.01)
CPC G06T 17/20 (2013.01) [G06F 30/17 (2020.01); G06F 30/23 (2020.01); G06T 5/002 (2013.01)] 11 Claims
OG exemplary drawing
 
1. A computer-implemented method for 3D reconstruction, the method comprising:
obtaining 2D images which all represent a same real object which is fixed, and, for each 2D image, camera parameters which define a perspective projection;
obtaining, for each 2D image, a smooth map, the smooth map having pixel values each representing a measurement of contour presence, the contour being a location of separation between two different textures of the 2D image; and
determining a 3D modeled object that represents the real object, the determining iteratively optimizing energy that rewards, for each smooth map, projections of silhouette vertices of the 3D modeled object having pixel values, on the smooth map, representing a high measurement of contour presence,
wherein the energy includes a term of a type:

OG Complex Work Unit Math
where for a 2D image i:
Σi(⋅) denotes the smooth map,
sij denotes a silhouette point j, and
πi(⋅) denotes a projection transformation.