Page 70 - Video Coding for Mobile Communications Efficiency, Complexity, and Resilience
P. 70

Section 3.3.  Chronological  Development                       47


            comparisons indicate that H.263 provides the same subjective quality as H.261
            but with less than half the bit rate [65]. This performance improvement is due
            to  optimized  coding  techniques  as  well  as  advanced  optional  coding  modes.
            Some  of  the  new  features  of  H.263  compared  to  H.261  are  the  support  for
            more picture formats, half-pel MC, a 3-D (LAST-RUN-LEVEL) RLE instead
            of  2-D  (RUN-LEVEL),  more  optimized  VLC  tables,  optional  extra  head-
            ers  to  increase  error  resilience,  advanced  2-D  median  predictor  for  motion
            vector  coding,  more  optimized  macroblock  addressing  and  quantization  adap-
            tation, optional extended-range unrestricted motion vectors that can point out-
            side  frames,  optional  arithmetic  coding,  optional  advanced  prediction  with
            overlapped  motion  compensation  and  four  motion  vectors  per  macroblock,
            and  optional  bidirectional  prediction.  H.263  is  described  in  more  detail  in
            Section 3.4.


            3.3.8  H.263+

            Technically, H.263+ is version 2 of the H.263 standard [66]. This version was
            developed  by  ITU-T=SG16=Q15  Advanced  Video  Experts  Group  (previously
            under  ITU-T=SG15),  with  technical  content  completed  in  1997  and  approved
            in 1998. The H.263+ standard added 12 new optional features to H.263. These
            new  features  support  custom  picture  size  and  clock  frequency,  improve  com-
            pression  e,ciency,  allow  scalability,  enhance  error  resilience  over  wireless
            and  packet-based  networks,  provide  supplemental  display  and  external  usage
            capabilities, and ensure backward compatibility. H.263+ is described in more
            detail in Section 3.4.


            3.3.9  MPEG-4

            In  1993,  the  ISO=IEC  JTC1=SC29=WG11  MPEG  group  initiated  a  new  stan-
            dardization activity called MPEG-4. The target was the very-low-bit range and
            the aim was to achieve higher compression e,ciency than could be achieved
            by  existing  conventional  techniques.  In  1994,  it  was  realized  that  too  few
            improvements  could  be  achieved  over  the  H.263  and  H.263+  compression
            results  to  justify  a  new  standard.  Thus,  the  group  decided  to  broaden  the  ob-
            jectives  of  the  MPEG-4  e1ort  and  started  an  in-depth  analysis  of  the  trends
            within the audiovisual world. Particular attention was given to the convergence
            of  the  three  traditionally  separate  industries  of  communications,  computing,
            and  TV=/lm=entertainment.  This  study  concluded  that  MPEG-4  should  sup-
            port  functionalities  that  would  be  useful  in  future  applications  but  were  not
            supported  or  not  well  supported  by  the  available  standards.  Eight  main  new
            or improved functionalities were identi/ed and then clustered in three classes:
   65   66   67   68   69   70   71   72   73   74   75