| With the rapid development of computer and digital communication, image compression is becoming a research focus since the bandwidth and capability of communications is limited. As an important entropy coding method, arithmetic coding has excellent performance;its average code length can always approach the information source's entropy, and its related research is concerned increasingly. Arithmetic coding is adopted by many international standards as essential element.JPEG2000 is the next generation digital still image compression standard introduced by Joint Photographic Experts Group in December, 2000. JPEG2000 provides not only superior compression performance over JPEG, but also a rich set of features, in answer to the ever growing requirements for image coding techniques. MQ encoder is adopted as JPEG2000's entropy method.In this thesis, theory of arithmetic coding is researched carefully, and some important concept about binary arithmetic coding is discussed, such as multiple-free, condition exchange and adaptive probability estimation. By optimizing algorithms and analyzing the details of MQ encoding, a three pipelined architecture of MQ encoder which can operate one symbol per clock cycle is introduced. In order to reach higher throughput's requirement, parallel encoding two symbols synchronously is discussed. By analyzing the design's difficulties and giving corresponding solving strategy, a four pipelined stage's architecture is implemented finally.The given architecture is described with Verilog HDL in register transfer level (RTL), and simulated to verify the functional correction. By compared and analyzed the performances, the first introduced architecture can work at 200MHz, consumed 7486 logic gate counts;while the second given architecture can work at 143MHz, consumed 12547 logic gate counts. The second architecture can operate 47×10~6 pixels/s for grayscale image, which meets the higher throughput's demands to match the parallel bit plane encoder better. |
PDF Full Text Request