if any one knows about the N-LEVEL QUANTIZATION IN HEVC here the link is
in hm6.0 where we have to satrt writing the code for the above paper
From what I've read of H.265, it is an evolutionary development of H.264, just as H.264 was an evolutionary development of MPEG-2 compression (H.262).
Usually, the efficiency increase figures that get bandied about don't reflect the actual reduction in bit rate, for whatever reasons. Although in the case of H.264, it does work much better than H.262 at very low bit rates, such as less than 2 Mb/s. But when H.264 is used for HD quality TV, the bit rates are not half of what they are when H.262 is used, as the hype might have suggested. Instead, they are comparable.
One problem with codec improvements is that they invariably require more processing power, both is coding and in decoding. The decoding part can become problematic for existing hardware, when the decoding is done in software. Still, if 3D TV or UHDTV are to become reality, there's no other choice but to keep improving codecs.
You don't have to gamble too much. Use a programmable SOC. Do your initial implementation and then update as the standard matures. Anyway these standards have a subjective image quality to begin with. You need to tweak your implementation over time.
An interesting gamble for chip designers. How early can they commit to silicon to be first to market ... while still ensuring that they have all the necessary functionality to enable any revisions to the standard to be implemented as software updates.
If the rest of the world stood still then we might get to that end point, but it doesn't. CES this year started showing 4K sets, and 3D is also popular. If we pull images fully into the 3rd dimension then we need essentially a solid modeling format. Whatever direction displays move CODECs will have to follow.
If these efficiency improvements are for real, could H.265 be the end of the line in video CODEC evolution? By the time H.265 is widely deployed, so too will LTE be widely deployed.
The large increase in wireless data rates combined with a huge reduction in video bit rate will greatly relieve the impact of video on network congestion -- video will be a lot more like any other data. At that point, will we ever need a more advanced CODEC?
Blog Doing Math in FPGAs Tom Burke 13 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...