الفهرس | Only 14 pages are availabe for public view |
Abstract It is more important to build video compression systems that give higher compression ratios with good isual quality in the same time. Beside the higher processing speed is an important factor in those compression systems. Those compression systems are used in high definition (HD) video devices. broadcasting more television channels on the satellites and mobile phones which can deal with the multimedia. The ideo compression operation depends on removing the temporal redundancies due to motions between the successive video frames. Then spatial redundancies within the frames an: rernox ed. Finally the entropy operation is appl ied to get the compression operat ion. The temporal redundancies are removed by applying Motion Estimation (ME) and Motion Compensations (Me) operations to get motion vectors and residual frames. The residual frames have lower energies than original frames so it is easier to be compressed. The spatial redundancies are removed by the transformations operations as di crete wavelets transform OWl. Finally the statistical redundancies are removed at the entropy stage. In the entropy stage the modem method called Embedded Block Coding with Optimized Truncation (EBCOT). used in the image compression standard JPEG 2000, is applied. The EBCOT method gives higher compression ratio with good visual quality. Designed MATLAB tool programs are used in the motion analysis and the transformation stages. Those programs are applied on standard video frames length 50 frames and of size 128xJ28 pixels. To ensure the accuracy of those programs the inverse of those programs are applied in the reverse direction. The original frames are compared with resulted frames and lose was small. The EBCOT MATLAB tool programs are applied on the OWT transformed residual frames form the motion and transformation stages. Then the compression ratio and the visual quality are calculated. The compression ratio is witb range 40% 10 75% with decreasing the visual quality between 34% to 60 % respectively. The main problem of the EBCOT algorithm is that it consists of t\VO complicated stages. The first stage called Context Formation in which the contexts decisions pairs are generated. The CF deals with bits of during frame processing so it needs longer processing period. The second stage is Arithmetic Coder which uses contexts decisions pairs to generate the compressed data. But also the AC is too complicated and need long processing period. The used Arithmetic Coder is famously known as Context Adaptive Arithmetic Coder (CABAC). To use the CABAC algorithm as an entropy method it is required to design high speed architecture to overcome its long processing period This thesis presents high speed design for the EBCOT algorithm on field programmable gate array (FPGI\). So the proposed design can deal with the digital video signals. The simulation is done on chip of type Yirtex xcv200e of Xillinx Company. This chip has a feature high speed performance and high capacity. The simulation results indicate that the design can operate at frequency 93 MHz after place and routing and can process a video frame of size 128x 128 pixels in 7.8 m St:C. Also from the comparing the proposed design with some modern designs, this design achieves higher processing rate with 51 % and decreasing the deJay time with 56.6 % but with increase in the used gate with 37% and more power consumption with about 24%. |