Microsoft Technology Licensing, LLC

Ali Osman Ulusoy

Yuri Pekelny

Szymon Piotr Stachniak

Michelle Brook

Hendrik Mark Langerak

Salah Eddine Nouri

The described implementations relate to processing of an environment using a plurality of sub-volumes, and specifically to generating surface representations in the plurality of sub-volumes for individual processing. One example can identify planar fragments within the plurality of sub-volumes. The example can determine that various planar fragments constitute part of a contiguous planar surface and should be aggregated. The example can also output data representing the contiguous planar surface formed from the aggregated planar fragments.

Planar surface detection

A computer system is provided that includes a camera device and a processor configured to receive scene data captured by the camera device for a three-dimensional environment that includes one or more physical objects, generate a geometric representation of the scene data, process the scene data using an artificial intelligence machine learning model that outputs object boundary data and object labels, augment the geometric representation with the object boundary data and the object labels, and identify the one or more physical objects based on the augmented geometric representation of the three-dimensional environment. For each identified physical object, the processor is configured to generate an associated virtual object that is fit to one or more geometric characteristics of that identified physical object. The processor is further configured to track each identified physical object and associated virtual object across successive updates to the scene data.

Object permanence in surface reconstruction

A method for spatial mapping of an environment comprises receiving a plurality of depth images of an environment via an imaging device, each depth image associated with a local coordinate system. For each local coordinate system, each associated depth image is fused to generate a local volume. Each local volume is then fused into a global volume having a global coordinate system, and then a surface mesh is extracted for the global volume. One or more regions of inconsistency within the global volume are determined and localized to one or more erroneous local volumes. The one or more erroneous local volumes are unfused from the global volume, and then non-erroneous local volumes are re-fused into a corrected global volume. By using a two-step fusion process, regions of inconsistency, such as mirror reflections, may be corrected without requiring reconstruction of the entire global volume.

Fusion of depth images into global volumes

A computing system is provided, including one or more optical sensors, a display, one or more user input devices, and a processor. The processor may receive optical data of a physical environment. Based on the optical data, the processor may generate a three-dimensional representation of the physical environment. For at least one target region of the physical environment, the processor may generate a three-dimensional bounding volume surrounding the target region based on a depth profile measured by the one or more optical sensors and/or estimated by the processor. The processor may generate a two-dimensional bounding shape at least in part by projecting the three-dimensional bounding volume onto an imaging surface of an optical sensor. The processor may output an image of the physical environment and the two-dimensional bounding shape for display. The processor may receive a user input and modify the two-dimensional bounding shape based on the user input.

Growing community of inventors

Seattle, WA, United States of America

Ali Osman Ulusoy