Video encoders

Deinterlacing techniques

At left, a close-up of an interlaced image shown on a computer screen; at right, the same interlaced image with deinterlacing technique applied.

Video from analog cameras is designed to be viewed on analog monitors such as traditional TV sets, which use a technique called interlaced scanning. With interlaced scanning, two consecutive interlaced fields of lines are shown to form an image.

When such video is shown on a computer screen, which uses a different technique called progressive scanning, interlacing effects (i.e., tearing or comb effect) from moving objects can be seen. In order to reduce the unwanted interlacing effects, different deinterlacing techniques can be employed. In advanced Axis video encoders, users can choose between two different deinterlacing techniques: adaptive interpolation and blending.

Adaptive interpolation

Adaptive interpolation offers the best image quality. The technique involves using only one of the two consecutive fields and using interpolation to create the other field of lines to form a full image.

Blending

Blending involves merging two consecutive fields and displaying them as one image so that all fields are present. The image is then filtered to smooth out the motion artifacts or ‘comb effect’ caused by the fact that the two fields were captured at slightly different times. The blending technique is not as processor intensive as adaptive interpolation.

Next topic: Video decoder

Video decoder