To achieve this, the Inventors introduce a specific camera movement during exposure. This movement is designed so that the compound motion of the camera and any object velocity along the selected orientation and at any depth results in the same easy-to-invert PSF. Since the entire scene is blurred with an identical PSF, a single deconvolution kernel can be used to remove blur and create sharp images of scenes with objects moving at different speeds, without requiring any segmentation, and without knowledge of the object speeds. In practice, motions close to the selected 1D orientation can also be deblurred.
This approach is inspired by wavefront coding, where depth of field is improved by modifying a lens to make the defocus blur invariant to depth and easy to invert. When motion blur is analyzed as the integration over curves resulting from camera and object motion, the only integration curve that results in a motion-invariant PSF is a parabola. This corresponds to constant 1D acceleration of the camera, which first moves quickly in one direction, progressively slows down until it stops, then picks up speed in the other direction. The observation that for any object velocity within a range, there is always one moment during exposure where the camera is perfectly tracking the object, motivates the camera movement in this technology.