Automatic spatial layout determination and estimation

A non-transitory computer readable medium (CRM) comprising instructions that, when executed by an apparatus, cause the apparatus to:

• capture, with a capturing device, depth information of a three-dimensional (3D) space;
• generate, from the depth information, a top-down 3D representation of the 3D space;
• assist a user of the capturing device to complete a closed loop scan for layout estimation of the 3D space by:
• identify, from the 3D representation, a plurality of planes;
• identify, from the plurality of planes, a floor plane, the floor plane being a plane with a largest area that is located most distant from a top-down camera view of the 3D representation;
• collapse depth information that is above the floor plane onto the floor plane to obtain a 2D image, where the 2D image reflects the collapsed depth information;
• detect one or more contours in the 2D image;
• identify one or more segments located in the one or more detected contours;
• create, from a contour with a largest area detected in the one or more contours, a 2D layout with a scale that is identical to the 3D space, the 2D layout including the one or more segments localized onto the contour;
• and
• display the 2D layout on a display device to the user to enable the user to adjust a location of at least a portion of the one or more segments or to add and subtract from the one or more segments.

A system, comprising:

• a server with a processor, and
• a storage device in communication with the server,
• wherein the storage device includes instructions that, when executed by the processor, cause the server to:
• receive, from a capturing device, a video and correlated depth data of a three-dimensional (3D) space;
• generate, from video and motion data, a 3D mesh representation of the 3D space;
• assisting a user of the capturing device to complete a closed loop scan for layout estimation of the 3D space by:
• orient, using the motion data, the 3D mesh representation into a top-down camera view;
• identify, from the 3D mesh representation, one or more planes;
• assign, from the one or more planes, a plane with a largest area that is located most distant from the top-down camera view as a floor plane;
• identify, from the 3D mesh representation, an exterior boundary of the floor plane;
• identify, from the 3D mesh representation and motion data, one or more segments of the exterior boundary located in at least one of the one or more planes;
• determine, from the 3D mesh representation, the floor plane, and one or more identified planes, a 3D physical layout of each of the one or more planes that is of equal scale to the 3D space, the 3D physical layout including the one or more segments localized onto the at least one of the one or more planes;
• and
• display the 3D physical layout on a display device to the user to enable the user to adjust a location of at least a portion of the one or more segments of the exterior boundary or to add and subtract from the one or more segments.