Method and apparatus to correct digital image blur by combining multiple images

What is this patent about?

’587 is related to the field of digital image processing, specifically addressing the problem of image blur. In traditional photography, relative motion between the camera and the subject during a long exposure results in a smudged representation of light across the sensor. While hardware-based stabilization and software-based sharpening exist, they often introduce weight, cost, or data loss, failing to truly recover the original, sharp image data lost during the motion event.

The underlying idea behind ’587 is to replace a single, long-exposure capture with a sequence of rapid-fire, high-speed captures that are intelligently reconstructed. By operating the sensor at its maximum shutter speed to take multiple back-to-back images, the system ensures that each individual frame is sharp and free of motion displacement. These frames are then mathematically aligned and merged to synthesize the equivalent of a long exposure, maintaining high signal quality without the associated motion artifacts.

The claims of ’587 focus on a specific workflow for subject-based stabilization involving a viewfinder and a processor. The method requires displaying a live image, designating a main subject within that field of view—either automatically by a processor or via user input—and capturing a plurality of images containing that subject. The system then generates a corrected image by calculating new pixel values based on the corresponding image points across the entire captured sequence.

In practice, the invention functions by treating the designated subject as the anchor for the reconstruction process. Once the subject is identified, the processor aligns the multiple high-speed frames so that the subject remains stationary across the stack, even if the camera or the subject moved during the sequence. By combining these aligned pixels, the system performs a coherent addition of image data, which effectively boosts the signal-to-noise ratio and overcomes the low-light limitations typically associated with fast shutter speeds.

This approach differs from prior solutions by moving away from mechanical lens compensation or destructive post-capture sharpening. Instead of trying to prevent motion physically, it embraces the motion and uses temporal oversampling to reconstruct a sharp result. By focusing the alignment on a specific designated subject, the invention allows for selective stabilization, ensuring that the most important part of the composition remains crisp even when the surrounding environment or the photographer is in motion.