| No. 0, July 2003 - Art. 2 |
| |
| An Innovative Adaptive Noise Reduction Filter for Moving Pictures Based on Modified Duncan Range Test |
|
 |
by
Antonio Borneo, Lanfranco Salinari, Daniele Sirtori (STMicroelectronics)
Copyright
© STMicroelectronics, 2000
|
|
| |
Abstract
To improve the perceived video quality of TV, every high-end set requires noise reduction, which is also a key preprocessing task for consumer digital video compression, such as MPEG2.
This paper presents an innovative noise reduction method targeted for TV video sequences.
It is based on the DRT (Duncan Range Test) statistical test, properly modified to meet both image processing and real time requirements.
A clever integration of the above filter in the MPEG2 encoder system architecture allows both filtering and coding improvement because of its proprietary Motion Estimator structure, which also allows for the implementation of an innovative method of noise variance measurement. |
| |
|
| |
|
|
|
