In applications where digital images must be transmitted over a slow
channel and displayed on a reasonably fast computer system, the use of
foveation can be of great benefit in reducing transmission times.
These benefits can be realized without the aid of an eye-tracking
device, although the resultant blur will be noticeable to the
observer. Compression ratios of over 80% have been achieved for
foveation of large images, and even greater savings could be imagined
if the foveation function were allowed to decrease more rapidly than
the human visual system. If desired, foveation could be increased
until a desired compression ratio is achieved for a given image.
If the blurring due to foveation is not acceptable in a final image,
but rapid image recognition is desired, then a fovea-first
transmission method might be appropriate. In a medical imaging
context, imagine a radiologist looking through hundreds of MRI scans
for a patient who has a particular ailment. If each image were
foveated with the point of gaze centered on the patient's ailment,
then the radiologist could rapidly scroll through many images. If the
image currently being rendered does not match the radiologist's search
criteria, then he or she could move on before the entire image is
transmitted. However, when the radiologist finds the patient that he
or she is looking for, then the entire high-resolution image will
eventually be received.
These methods of image compression and progressive image
transmission have certain niche applications where they could be
marginally useful. Mainstream image compression standards are not
likely to change anytime soon in order to accomodate these specialized
situations, but there are purposes for which their implementation may
well be worth the additional complexity.