Author :
Aeluri, Praveen K. ; Bojan, Vinodh ; Richie, Samuel ; Weeks, Arthur
Abstract :
Video quality estimation methods in the literature can be broadly classified into two classes: objective and subjective. The metrics obtained from both classes of methods represent different intrinsic features of a video (Eude, T. and Mayache, A., Proc. ICSP´98, 1998). However, we have performed an objective quality evaluation on various videos, which range from no motion to high motion, encoded by three encoders (MPEG-1, MPEG-2 and Windows Media formats) at two different frame rates, two frame sizes and three low bit rate combinations for the purpose of determining the change in video content introduced by the encoding and decoding process. The effect of the encoding, bit rates, frame rates and frame sizes on the quality of videos has been studied. We have defined and used a quality metric, cumulative brightness error. It is the cumulative mean square error of each frame´s color components, averaged for the entire video sequence. It enables comparison of video formats based on content differences and encoding parameter values.
Keywords :
data compression; decoding; mean square error methods; parameter estimation; video coding; MPEG-1; MPEG-2; Windows Media; bit rates; cumulative brightness error; decoding; encoding; frame rates; frame sizes; mean square error; objective video quality analysis; video compression; Bit rate; Brightness; Decoding; Encoding; Humans; Image coding; Performance evaluation; Streaming media; Transform coding; Video compression;