Ray-Based Occlusion Detection with Real-World Space Models

The patent introduces a digital model as an intermediary between the physical environment and the spatial computing system. This digital model represents the real-world space and contains information about occluding features, enabling the system to detect hidden objects without directly observing them. The digital model acts as a mediator that bridges the gap between visible and hidden spatial information.

The patent creates a digital copy or representation of the real-world space through a digital model. This copy contains simplified geometric representations of occluding features and spatial relationships, allowing the system to analyze occlusions computationally without requiring direct sensory input for every object. The digital twin concept enables inference about hidden objects based on the copied spatial structure.

The patent segments the occlusion detection problem into distinct computational steps: transforming real-world coordinates to digital model coordinates, generating rays from the camera position, intersecting rays with the digital model, and determining occlusion status. This segmentation allows each step to be optimized independently and facilitates efficient processing by breaking down the complex detection task into manageable operations.

The patent replaces physical or mechanical occlusion detection methods with computational geometry algorithms. Instead of using multiple physical sensors or cameras to detect occlusions, the system uses mathematical ray-tracing algorithms against a digital model. This substitution reduces hardware complexity while maintaining detection accuracy through efficient computational methods.