A technique for increasing the frame rate of CMOS video cameras is presented. The method uses the non-destructive readout capabilities of CMOS imagers to obtain low-speed, high-resolution frames and high-speed, low-resolution frames simultaneously. The algorithm translates the pixels of the full resolution images with respect to the motion dynamics observed in the low-resolution frames and corrects the result as necessary for consistency with the low-resolution frames. Noting that due to the longer exposure time required, high-resolution frames are more prone to motion blur than low-resolution frames, and thus, a motion blur reduction step is also applied. Simulations demonstrate the ability of our technique in synthesizing high-quality, high-resolution frames at modest computational expense.