Modeling indoor scenes using measurements captured using mobile devices

A method comprising:

• receiving a plurality of digital images depicting an indoor scene of a room and captured using a capture device;
• receiving trajectory data of a trajectory that the capture device followed as the capture device was capturing the plurality of digital images;
• wherein the plurality of digital images and the trajectory data satisfy a plurality of quality checks, including a camera inward facing check;
• wherein the camera inward facing check specifies that the capture device is to point inward the room so that a yaw angle of a direction of the capture device is facing toward a center of the room;
• wherein an inward-facing quality check method uses the trajectory data of the trajectory to calculate a percentage of images, from the plurality of digital images, for which the capture device used to capture the plurality of digital images was inward-facing;
• determining a point cloud by preprocessing the plurality of digital images and the trajectory data to satisfy a plurality of optimization constraints, and by identifying and labelling a plurality of feature points in data obtained by preprocessing the plurality of digital images and the trajectory data;
• based on the point cloud and the trajectory data, generating room geometry for the room;
• based on the room geometry, generating a 3D model of the room;
• generating a graphical representation of the 3D model and displaying the graphical representation of the 3D model on a computer display device.

A non-transitory computer readable storage medium storing instructions which, when executed by one or more computing devices, cause the one or more computing devices to perform:

• receiving a plurality of digital images depicting an indoor scene of a room and captured using a capture device;
• receiving trajectory data of a trajectory that the capture device followed as the capture device was capturing the plurality of digital images;
• wherein the plurality of digital images and the trajectory data satisfy a plurality of quality checks, including a camera inward facing check;
• wherein the camera inward facing check specifies that the capture device is to point inward the room so that a yaw angle of a direction of the capture device is facing toward a center of the room;
• wherein an inward-facing quality check method uses the trajectory data of the trajectory to calculate a percentage of images, from the plurality of digital images, for which the capture device used to capture the plurality of digital images was inward-facing;
• determining a point cloud by preprocessing the plurality of digital images and the trajectory data to satisfy a plurality of optimization constraints, and by identifying and labelling a plurality of feature points in data obtained by preprocessing the plurality of digital images and the trajectory data;
• based on the point cloud and the trajectory data, generating room geometry for the room;
• based on the room geometry, generating a 3D model of the room;
• generating a graphical representation of the 3D model and displaying the graphical representation of the 3D model on a computer display device.