Image analysis – Pattern recognition – Feature extraction
Reexamination Certificate
1998-09-10
2004-09-21
Mehta, Bhavesh M. (Department: 2621)
Image analysis
Pattern recognition
Feature extraction
C348S014120, C348S092000, C348S193000, C348S625000, C382S112000, C382S266000, C382S274000, C455S135000, C455S226300, C702S069000, C702S081000
Reexamination Certificate
active
06795580
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to picture quality measurement, and more particularly to picture quality measurement using blockiness.
In digital video transmission where bandwidth is important, such as transmission over satellite link, the video signal is subjected to varying degrees of compression to decrease the bandwidth required for each video channel. The compression standards typically used, such as JPEG, MPEG or proprietary variants thereof, are “lossy”—to achieve higher compression they allow distortions to occur in an image represented by the video signal. The amount of distortion is a function of the complexity of the image and of the number of bits per second (bit rate) a compression encoder is allowed to use. Ideally the amount of compression is maximized while still providing a video image or picture to the customer that is free of disturbing artifacts.
Current devices to analyze picture quality, such as the PQA200 Picture Quality Analyzer manufactured by Tektronix, Inc. of Wilsonville, Oreg., USA, are reference-based. A video signal transmitted through a video system is compared with the original video signal as a reference video signal in a measurement device. The reference video signal is either stored in the measurement device or is transmitted via some other non-distorted pathway to the measurement device. Use of a reference video signal is necessary for extremely accurate algorithms, such as the Sarnoff Corporation JNDmetrix™ human visual model algorithm. However, this means that measurements are only made on video signals whose contents are either known in advance or are immediately available, such as double-checking an encoder at the source.
Other potential methods of measuring discrete cosine transform (DCI) based codec degradations involve directly examining the coarseness of the quantization scale in the compressed video stream, optionally combined with a measure of the complexity of the original image transmitted by some means outside the video channel—a form of compressed reference. This method is not as accurate, and in any case can only make measurements on compressed video, not video that has been already decompressed and potentially passed through other systems, including other additional codecs, prior to end-user delivery.
What is desired is a method of monitoring distortions in block-based codecs in a manner that is truly reference-less and which uses the minimum amount of hardware acceleration to perform such measurements in real time.
BRIEF SUMMARY OF THE INVENTION
Accordingly the present invention provides a picture quality measurement system using blockiness to determine the amount of degradation of a video signal due to video compression processing. A frame/field of a processed video signal representing an image or picture is captured and converted, if necessary, into luminance and color components. One or more of the components is analyzed by appropriate vertical and/or horizontal edge enhancement filtering, such as boxcar filtering. The resulting edges are correlated with an infinite grid having boundaries corresponding to the block boundaries used in the video compression. Optionally a second correlation may be made with a grid with boundaries slightly different than the block boundaries used in the video compression, with this result being subtracted from the first value. Further the locations of the compression block boundaries may be detected by observing where the maximum correlation value occurs. The resulting correlation results are processed to generate a picture quality rating for the image which represents the amount of human-perceivable block degradation that has been introduced into the processed video signal.
The objects, advantages and other novel features of the present invention are apparent from the following detailed description when read in conjunction with the appended claims and attached drawing.
REFERENCES:
patent: 4246608 (1981-01-01), Baker
patent: 4839727 (1989-06-01), Tanaka et al.
patent: 5410356 (1995-04-01), Kikuchi et al.
patent: 5446492 (1995-08-01), Wolf
patent: 5479215 (1995-12-01), Chmielewski et al.
patent: 5541668 (1996-07-01), Kuwahara et al.
patent: 5550937 (1996-08-01), Bell et al.
patent: 5583579 (1996-12-01), Shim
patent: 5661527 (1997-08-01), Ferguson
patent: 5748231 (1998-05-01), Park et al.
patent: 5796875 (1998-08-01), Read
patent: 5832101 (1998-11-01), Hwang et al.
patent: 5844614 (1998-12-01), Chong et al.
patent: 5886745 (1999-03-01), Muraji et al.
patent: 5905541 (1999-05-01), Yun
patent: 5920356 (1999-07-01), Gupta et al.
patent: 5933543 (1999-08-01), Cok
patent: 5963677 (1999-10-01), Schrick et al.
patent: 6018361 (2000-01-01), Fujii et al.
patent: 6038339 (2000-03-01), Hubel et al.
patent: 6075884 (2000-06-01), Lubin et al.
patent: 6078689 (2000-06-01), Kunitake et al.
patent: 6097848 (2000-08-01), Salvati
patent: 6101222 (2000-08-01), Dorricott
patent: 6137824 (2000-10-01), Liu
patent: 6154577 (2000-11-01), Warnick et al.
patent: 6239834 (2001-05-01), Miyaji et al.
patent: 6272253 (2001-08-01), Bannon et al.
patent: 6408109 (2002-06-01), Silver et al.
patent: 6418242 (2002-07-01), Maurer
patent: 6434275 (2002-08-01), Fukuda et al.
patent: 6535254 (2003-03-01), Olsson et al.
patent: 6594595 (2003-07-01), Yamaguchi et al.
G. Beakley, C. Cressy, and J. Van Pelt, “Video Compression: The Need for Testing” SMPTE Journal, Nov. 1995, Vo. 104, No. 11, pp. 742-750.
Janko Bozidar
Maurer Steven D.
Desire Gregory
Gray Francis I.
Mehta Bhavesh M.
Tektronix Inc.
LandOfFree
Picture quality measurement using blockiness does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Picture quality measurement using blockiness, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Picture quality measurement using blockiness will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3244361