Deinterlacing with source, motion and edge-adaptive processing reduces artifacts on HDTVs

The reality is that their new HDTV is capable of displaying all the details in the output image, including noise and image artifacts, which their lower-resolution display was previously masking. Some of these artifacts are caused by poor deinterlacing, a technology required to convert standard-definition interlaced signals such as NTSC and PAL to a progressive-scan signal that can be displayed on a digital display. Other artifacts are caused by digital image compression, a technology required to transmit hundreds of channels with limited bandwidth.

Interlace basics
Most TV channels that are available through digital satellite and digital cable are older standard-definition signals. All standard-definition content is captured and broadcast in a method called interlace.

An interlaced video signal is captured and sent in two alternate fields. The first field drawn on the screen is all the odd lines. This is followed by the second field, made up of the even lines. The odd lines are drawn in between the even lines, hence the term interlaced. By taking advantage of the persistence of vision effect, a theory that states the perceptual processes of the retina of the human eye retain an image for a brief moment, this method is sufficient in creating the illusion of a higher-resolution image (Figure 1).

View full size

Figure 1: These four images are actually taken from a sequence of four video fields. The flag is in motion, but you have to look closely to see it, because the total elapsed time from the left image to the right one is 3/60 sec, or about 50 milliseconds. The motion is most visible in the small corner of the flag that hangs down at the bottom.

Next: Motion Adaptive Deinterlacing