Floorplan generation based on room scanning

1. A method comprising:

at an electronic device having a processor:

displaying a live camera feed comprising a sequence of images of a physical environment;

obtaining a three-dimensional (3D) representation of the physical environment that was generated based on depth data and light intensity image data of the physical environment obtained during the displaying of the live camera feed;

generating a live preview of a preliminary two-dimensional (2D) floorplan of the physical environment based on the 3D representation of the physical environment utilizing a first process;

displaying the live preview of the preliminary 2D floorplan concurrently with the live camera feed; and

generating, based on the 3D representation, a final 2D floorplan of the physical environment utilizing a second process that is different than the first process, wherein the second process includes:

classifying features based on the 3D representation;

generating a transitional 2D floorplan based on the classified features;

determining refinements for the transitional 2D floorplan; and

generating the final 2D floorplan of the physical environment based on the determined refinements for the transitional 2D floorplan.

2. The method of claim 1 , wherein classifying the features based on the 3D representation comprises classifying corners and small walls based on the 3D representation using a more computationally intensive neural network.

3. The method of claim 1 , wherein generating the live preview of the 2D floorplan comprises:

generating 2D semantic data for multiple horizontal layers of the physical environment based on the 3D representation; and

generating the 2D floorplan using the 2D semantic data.

4. The method of claim 1 , wherein the 3D representation is associated with 3D semantic data that includes a 3D point cloud that includes semantic labels associated with at least a portion of 3D points within the 3D point cloud.

5. The method of claim 4 , wherein the semantic labels identify walls, wall attributes, objects, and classifications of the objects of the physical environment.

6. A device comprising:

a non-transitory computer-readable storage medium; and

one or more processors coupled to the non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium comprises program instructions that, when executed on the one or more processors, cause the device to perform operations comprising:

displaying a live camera feed comprising a sequence of images of a physical environment;

obtaining a three-dimensional (3D) representation of the physical environment that was generated based on depth data and light intensity image data of the physical environment obtained during the displaying of the live camera feed;

generating a live preview of a preliminary two-dimensional (2D) floorplan of the physical environment based on the 3D representation of the physical environment utilizing a first process;

displaying the live preview of the preliminary 2D floorplan concurrently with the live camera feed; and

generating, based on the 3D representation, a final 2D floorplan of the physical environment utilizing a second process that is different than the first process, wherein the second process includes:

classifying features based on the 3D representation;

generating a transitional 2D floorplan based on the classified features;

determining refinements for the transitional 2D floorplan; and

generating the final 2D floorplan of the physical environment based on the determined refinements for the transitional 2D floorplan.

7. The device of claim 6 , wherein classifying the features based on the 3D representation comprises classifying corners and small walls based on the 3D representation using a more computationally intensive neural network.

8. The device of claim 6 , wherein generating the live preview of the preliminary 2D floorplan comprises:

generating an edge map by identifying walls in the physical environment based on the 3D representation;

updating the edge map by identifying wall attributes in the physical environment based on the 3D representation;

updating the edge map by identifying objects in the physical environment based on the 3D representation; and

generating the live preview of the preliminary 2D floorplan based on the updated edge map that includes the identified walls, the identified wall attributes, and the identified objects.

9. The device of claim 8 , wherein generating the edge map by identifying walls further comprises:

determining parametrically refined lines for the edge map using a line fitting algorithm; and

updating the edge map based on the parametrically refined lines.

10. The device of claim 8 , wherein updating the edge map by identifying wall attributes comprises:

determining boundaries for the identified wall attributes using a wall attribute neural network and the sequence of images of the live camera feed; and

generating refined boundaries using a polygon heuristics algorithm based on the 3D representation associated with the identified wall attributes.

11. The device of claim 8 , wherein updating the edge map by identifying objects comprises:

generating 3D bounding boxes corresponding to the identified objects in the physical environment based on the 3D representation; and

generating 2D representations of the 3D bounding boxes.

12. A non-transitory computer-readable storage medium, storing program instructions executable on a device to perform operations comprising:

displaying a live camera feed comprising a sequence of images of a physical environment;

obtaining a three-dimensional (3D) representation of the physical environment that was generated based on depth data and light intensity image data of the physical environment obtained during the displaying of the live camera feed;

generating a live preview of a preliminary two-dimensional (2D) floorplan of the physical environment based on the 3D representation of the physical environment utilizing a first process;

displaying the live preview of the preliminary 2D floorplan concurrently with the live camera feed; and

generating, based on the 3D representation, a final 2D floorplan of the physical environment utilizing a second process that is different than the first process, wherein the second process includes:

classifying features based on the 3D representation;

generating a transitional 2D floorplan based on the classified features;

determining refinements for the transitional 2D floorplan; and

generating the final 2D floorplan of the physical environment based on the determined refinements for the transitional 2D floorplan.

13. The non-transitory computer-readable storage medium of claim 4 , wherein classifying the features based on the 3D representation comprises classifying corners and small walls based on the 3D representation using a more computationally intensive neural network.

14. A method comprising:

at an electronic device having a processor:

displaying a live camera feed comprising a sequence of images of a physical environment;

obtaining a three-dimensional (3D) representation of the physical environment that was generated based on depth data and light intensity image data of the physical environment obtained during the displaying of the live camera feed;

generating a live preview of a preliminary two-dimensional (2D) floorplan of the physical environment based on the 3D representation of the physical environment, wherein generating the live preview of the preliminary 2D floorplan comprises:

generating an edge map by identifying walls in the physical environment based on the 3D representation;

updating the edge map by identifying wall attributes in the physical environment based on the 3D representation;

updating the edge map by identifying objects in the physical environment based on the 3D representation; and

generating the live preview of the preliminary 2D floorplan based on the updated edge map that includes the identified walls, the identified wall attributes, and the identified objects; and

displaying the live preview of the preliminary 2D floorplan concurrently with the live camera feed.

15. The method of claim 14 , wherein generating the edge map by identifying walls further comprises:

determining parametrically refined lines for the edge map using a line fitting algorithm; and

updating the edge map based on the parametrically refined lines.

16. The method of claim 14 , wherein updating the edge map by identifying wall attributes comprises:

determining boundaries for the identified wall attributes using a wall attribute neural network and the sequence of images of the live camera feed; and

generating refined boundaries using a polygon heuristics algorithm based on the 3D representation associated with the identified wall attributes.

17. The method of claim 14 , wherein updating the edge map by identifying objects comprises:

generating 3D bounding boxes corresponding to the identified objects in the physical environment based on the 3D representation; and

generating 2D representations of the 3D bounding boxes.