Television – Monitoring – testing – or measuring
Reexamination Certificate
2001-12-11
2004-11-23
Lee, Michael H. (Department: 2614)
Television
Monitoring, testing, or measuring
C348S607000, C348S608000, C348S701000, C382S268000, C382S275000
Reexamination Certificate
active
06822675
ABSTRACT:
The present invention relates to a method of measuring video quality of an input sequence of digital pictures.
The invention also relates to a corresponding device for carrying out such a method of measuring video quality.
This invention may be used in, for example, digital television in order to create a measurement scale for impairments to digital video. As a consequence of said application, a method of post-processing may be applied to a sequence of pictures depending on the measurement scale of said sequence of pictures. Said post-processing method can reduce the visibility of visual impairments such as blocking artifacts, ringing artifacts and corner outliers.
Evaluation of video quality is generally performed using subjective tests, based on evaluation by naive human subjects. These methods are expensive, time consuming, subjective and not applicable for in-service testing, as they are based on comparison of an encoded sequence of pictures with an original sequence of pictures as reference.
Other methods have been developed to evaluate video quality. These methods are based on impairment metrics, and more especially on blocking artifact metrics. For example, such a method is described in the international application WO 01/20912. Said method relates to a method of detecting blocking artifacts in digital video pictures. The detection method comprises a step of filtering a digital input signal using a gradient filter for providing at least one filtered signal and a step of calculating a block level metric for processing the filtered signal(s) to identify and count blocking artifacts as a function of their position in a grid. If the block level metric is lower than a threshold, the picture has either not been encoded using a block-based processing, or has been encoded in a seamless way. In the opposite case, the picture has been encoded using a block-based processing in a non-seamless way and corrective actions, such as post-processing, can be taken.
But such a method is only able to measure a comparative blocking artifact level in order to apply a suitable post-processing, and, as a consequence, is neither able to perform a really objective measurement of the video quality, nor sensitive to other visual impairments.
It is an object of the invention to provide a method of measuring digital video quality for an input sequence of pictures, which provides a more objective quality metric.
To this end, the method of measuring digital video quality in accordance with the invention is characterized in that it comprises the steps of:
determining at least one reference level above which visual artifacts become noticeable to a group of subjects, with a corresponding predetermined artifact metric, from a set of reference sequences of digital pictures only comprising a corresponding artifact,
measuring at least one artifact level of the input sequence with the corresponding predetermined artifact metric,
computing a video quality metric of the input sequence from at least one ratio of the artifact level to the reference level corresponding to a same predetermined artifact metric.
Such a method of measuring video quality is based on the determination of at least one reference level. Thus, the amount of at least one predetermined visual artifact is weighted according to the at least one reference level. Said reference level can be seen as a perception factor and the determining step of said reference level as a perceptual calibration. As a consequence, the method of measuring video quality calculates a video quality metric with reference to at least one perceptual level, leading to a more objective quality metric.
Moreover, said method of measuring is multidimensional in the preferred embodiment, as it is based on a combination of impairment metrics. Said impairment metrics are determined from the three main visual artifacts resulting from digital compression, which are blocking, ringing and corner outlier artifacts, and which are independently quantified.
Thus, the method of measuring video quality is more complete compared with the prior art and leads to an even more objective quality metric.
REFERENCES:
patent: 4754492 (1988-06-01), Malvar
patent: 5454051 (1995-09-01), Smith
patent: 5495538 (1996-02-01), Fan
patent: 5654759 (1997-08-01), Augenbraun et al.
patent: 5748796 (1998-05-01), Pennino et al.
patent: 5799111 (1998-08-01), Guissin
patent: 5802218 (1998-09-01), Brailean
patent: 5819035 (1998-10-01), Devaney et al.
patent: 5844614 (1998-12-01), Chong et al.
patent: 5883983 (1999-03-01), Lee et al.
patent: 6061100 (2000-05-01), Ward et al.
patent: 6327307 (2001-12-01), Brailean et al.
patent: 6359658 (2002-03-01), He et al.
patent: 6600839 (2003-07-01), Mancuso et al.
patent: 6611295 (2003-08-01), Drouot et al.
patent: 0 881 837 (1998-02-01), None
patent: WO 97/37322 (1997-10-01), None
patent: WO0120912 (2001-03-01), None
“An Efficient Real Time Algorithm to Simultaneously Reduce Blocking and Ringing Artifacts of Compressed Video”, by Hong et al, Digital Media Research Lab., LG Electronics, Seoul Korea, 1999 IEEE, pp. 899-903.*
“Method for Eliminating Blocking Effect in Compressed Video Signal”, by Min-Cheol Hong, U.S. patent application Publication, US 2001/0003545 A1, Jun. 14, 2001.
Caviedes Jorge E.
Jung Jôel
Desir Jean W.
Koninklijke Philips Electronics , N.V.
Lee Michael H.
LandOfFree
Method of measuring digital video quality does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of measuring digital video quality, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of measuring digital video quality will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3296774