Multi-Sensor HDR Video Merging for Real-Time Capture

The patent divides the imaging function into multiple specialized sensors, each optimized for specific exposure ranges. Instead of using a single sensor to capture the entire dynamic range, the system segments the capture task across multiple sensors with different exposure settings, allowing each sensor to operate within its optimal range while collectively achieving high dynamic range capture.

The patent combines data from multiple sensors with different exposures through a merging process. The system integrates information from sensors capturing different exposure levels, selectively combining their outputs to reconstruct the full high dynamic range scene. This merging of multiple data streams enables the system to achieve dynamic range capabilities beyond what any single sensor can provide.

The patent performs preliminary actions by capturing multiple exposure levels simultaneously during the imaging process itself, rather than acquiring a single image and then processing it later. The multiple sensors capture different exposure ranges at the same moment, pre-preparing the data needed for HDR reconstruction before any post-processing occurs. This eliminates the time delay associated with sequential capture methods.

The patent establishes continuity of useful action by implementing a real-time processing pipeline that continuously merges data from multiple sensors as they stream in. Instead of batch processing complete frames, the system continuously integrates pixel data from multiple exposure streams, maintaining an uninterrupted HDR video output. This continuous processing eliminates idle time between capture and processing operations.