A new video noise reduction algorithm using spatial subbands

Noise filtering in television receivers provides an attractive feature, especially under sub-optimal reception conditions. This paper describes a concept of temporal recursive noise filtering in video signals for a standard interlaced TV environment based on the visual noise perception characteristics and on different processing modules in the spatial subbands. The subband (highs and lows channels) oriented algorithm allows optimal noise filtering based on a 2-channel-model of the human visual system. An adequate combination of two different subband based motion detection techniques is proposed. Therefore, different procedures in the highs and lows channels lead to results well matched to the requirements of the human visual system. An evaluation of the new concept is included. A comparison to known motion adaptive noise filter techniques shows a significant improvement of the scheme. Experiments using the new noise reduction scheme on MPEG-received video images show that a good quality improvement of noisy source materials can be achieved.