Temporal integration method for image-processing based super high resolution image acquisition

We present a temporal integration imaging method. This method uses a single camera with insufficiently low spatial resolution and/or low SNR, and produces an improved-resolution and/or improved-SNR image by integrating multiple consecutive image frames taken with the camera. However, it would seem that for the time being we cannot construct a universal temporal integration imaging method completely, because it would involve elaborating on image analysis techniques for segmentation and motion estimation of real moving image sequences far beyond the state of the art. We relinquish a generic approach and confine its applications in a particular type of application such as surveillance, where a user indicates a region of interest (ROI) on an observed low-resolution image before applying the temporal integration technique to an observed image sequence. The limitation of its application alleviates the requirements imposed on segmentation and motion estimation for temporal integration to a great extent, and hence we can easily construct an algorithm for integrating multiple image frames