ABSTRACT
Digital images and videos, acquired by still cameras, consumer camcorders, or even broadcast-quality video cameras, are usually degraded by some amount of blur and noise. In addition, most electronic cameras have limited spatial resolution determined by the characteristics of the sensor array. Common causes of blur are out of focus, relative motion, and atmospheric turbulence. Noise sources include film grain, thermal, electronic, and quantization noise. Further, many image sensors and media have known nonlinear input-output characteristics which can be represented as point nonlinearities. The goal of image and video (image sequence) restoration is to estimate each image (frame or field) as it would appear without any degradations, by first modeling the degradation process, and then applying an inverse procedure. This is distinct from the image enhancement techniques which are designed to manipulate an image in order to produce more pleasing results to an observer without making use of particular degradation models. On the other hand, superresolution refers to estimating an image at a resolution higher than that of the imaging sensor. Image sequence filtering (restoration and superresolution) becomes especially important when still images from video are desired. This is because the blur and noise can become rather objectionable when observing a ‘‘freeze-frame,’’ although they may not be visible
to the human eye at the usual frame rates. Since many video signals encountered in practice are interlaced, we address the cases of both progressive and interlaced video.
