US 12,283,007 B2
Method for generating an augmented image
Sascha Rau, Göppingen (DE); and Arnab Poddar, Esslingen (DE)
Assigned to TEAMVIEWER GERMANY GMBH, Göppingen (DE)
Filed by TeamViewer Germany GmbH, Göppingen (DE)
Filed on Mar. 16, 2023, as Appl. No. 18/122,517.
Claims priority of application No. 22163296 (EP), filed on Mar. 21, 2022.
Prior Publication US 2023/0298284 A1, Sep. 21, 2023
Int. Cl. G06T 19/00 (2011.01); G06V 10/74 (2022.01); H04N 5/265 (2006.01)
CPC G06T 19/006 (2013.01) [G06V 10/761 (2022.01); H04N 5/265 (2013.01); G06V 2201/07 (2022.01)] 10 Claims
OG exemplary drawing
 
1. A method implemented by a computer for generating an augmented image, the method comprising:
a) acquiring, from a first position, a first image of a three-dimensional (3D) scene using a camera;
b) obtaining 3D information of the 3D scene and calculating allocations of 3D coordinates to pixels in the first image;
c) detecting a user input representing a shape and a position of a spatially extended two-dimensional (2D) object in the first image;
d) calculating 3D coordinates of a point of interest that is associated with the 2D object, wherein this calculating step comprises
calculating a 3D centroid of a first 3D object that is derived from the 2D object;
e) calculating 3D coordinates of a second 3D object, thereby assuming that the second 3D object is located at the same distance from the camera as the point of interest;
f) calculating a first projection of the second 3D object, thereby using the 3D coordinates calculated in step e);
g) augmenting the first image by overlaying the first projection of the second 3D object calculated in step f);
h) moving the camera from the first position to a second position and acquiring, from the second position, a second image of the 3D scene;
i) calculating a second projection of the second 3D object for the second position, thereby using the 3D coordinates calculated in step e); and
j) augmenting the second image by overlaying the second projection of the second 3D object calculated in step i),
wherein in step d) the point of interest is calculated by
deriving the first 3D object from the 2D object by using the allocations of 3D coordinates to pixels in the image computed in step b),
calculating the 3D centroid for the first 3D object, and
deriving the point of interest from the 3D centroid, and
wherein the step of deriving the point of interest from the 3D centroid includes
projecting the 3D centroid on the first image,
calculating 2D coordinates of the 3D centroid in the first image,
considering pixels in a defined neighborhood of the 2D coordinates of the 3D centroid,
calculating distance values for the considered pixels, and
deriving the point of interest from the pixel for which the shortest distance to the camera is calculated.