Title :
An efficient no-reference blockiness metric for intra-coded video frames
Author :
Uzair, Muhammad ; Fayek, Dalia
Author_Institution :
Sch. of Eng., Univ. Of Guelph, Guelph, ON, Canada
Abstract :
When dealing with the quality of digital, block-based compressed and encoded video data, the objective measurement of the blocking artifacts plays an important role since “blockiness” in video sequences is deemed to be one of the most annoying sources of distortion. Objectively measuring the video quality enables the dynamic monitoring and the parameter adjustment in a variety of image and video processing applications. In this paper, we propose a No-Reference (NR) metric to measure the blocking distortions by using the Singular Value Decomposition (SVD) method. The metric measures the SVD values along the block boundaries in both the horizontal and vertical directions. It also incorporates the pixels that are near the boundaries. The difference of the SVD values gives a prediction of the blocking artifacts. The proposed metric is simple and provides computational savings. Our results show that our metric has a strong correlation with the subjective quality scores.
Keywords :
singular value decomposition; video coding; block-based compressed data; blocking distortions; dynamic monitoring; encoded video data; image processing; intracoded video frames; no-reference blockiness metric; parameter adjustment; singular value decomposition method; video processing; video sequences; Databases; Distortion measurement; Image coding; Image quality; Streaming media; No-Reference; blockiness; objective video quality measurement; singular value decomposition;
Conference_Titel :
Wireless Personal Multimedia Communications (WPMC), 2011 14th International Symposium on
Conference_Location :
Brest
Print_ISBN :
978-1-4577-1786-4
Electronic_ISBN :
1347-6890