Reduced Pixel Fill Factor in Projection Displays Using Shift-Add Fusion

The patent introduces a temporal dimension to the display system by implementing shift-add fusion that projects multiple shifted images sequentially at different time points. This temporal multiplexing allows the system to achieve higher effective resolution without reducing physical pixel size, as each pixel contributes to multiple shifted image positions across different time frames. The approach transforms a spatial resolution problem into a temporal processing solution.

The patent segments the image projection process into multiple discrete shifted image projections. Instead of projecting a single static image, the system divides the frame into multiple shifted versions that are projected sequentially. Each shifted image is a segment of the overall display task, and their combination through deconvolution algorithms achieves super-resolution without requiring smaller physical pixels.

The patent implements dynamic image shifting by tilting the illumination lens at an offset angle greater than twice the micromirror tilt angle. This dynamic optical configuration enables the projection of multiple shifted images without requiring additional physical pixels or expanding the pixel grid area. The dynamic tilting mechanism allows the same pixel array to generate multiple effective image positions through temporal multiplexing.

The patent changes the optical parameters of the illumination system by adjusting the incident angle of light on the micromirrors to be greater than twice the micromirror tilt angle. This parameter change enables the micromirrors to redirect light to multiple different angles, creating multiple shifted image projections from the same pixel array without requiring additional pixels or increased grid area.

The patent implements a feedback mechanism through deconvolution algorithms that process the multiple shifted images to reconstruct a sharp high-resolution image. The system captures multiple shifted image projections and uses computational feedback to eliminate overlap and blurring effects. This feedback processing restores image sharpness and achieves super-resolution by compensating for the degradation introduced during the shift-add fusion process.

The patent introduces deconvolution algorithms as an intermediary processing step between the shifted image projections and the final displayed image. This computational intermediary separates and reconstructs the overlapping shifted images, removing blurring effects and producing a sharp high-resolution output. The intermediary processing layer transforms the degraded shifted images into a quality high-resolution image without requiring additional hardware.