Methods for markerless tracking and latency reduction, and related devices.

Abstract

A method for tracking at least one object in an environment includes a generation step of generating a point cloud from image data acquired from cameras, the generation step of which includes a substep of imaging the at least one object and having each camera provide a corresponding image stream, and a substep of generating the point cloud as a three-dimensional representation of the at least one object based on the image stream; a segmentation step of segmenting the generated point cloud into a segmented object model for the at least one object, the segmentation step of which the segmentation identifies the at least one object and the boundaries of the at least one object as shown by the generated point cloud; and a deformation step of which a predefined mesh model representing at least a portion of the at least one object is deformed to correlate with the segmented object model, the deformation step of which the deformation provides the location of the at least one object in the environment.

Images