FALL EE 5359 2010 MULTIMEDIA PROCESSING PROJECT REPORT Performance Analysis of Dirac Pro with H 264 Intra frame coding INSTRUCTOR DR K R RAO Poonam S Kharwandikar Department of Electrical Engineering University of Texas at Arlington Email poonam kharwandikar mavs uta edu ABSTRACT Dirac Pro is a hybrid motion compensated state of the art video codec that uses modern techniques such as wavelet transforms and only Intra frame coding It is an open technology designed to avoid patent infringement and can be used without the payment of license fees H 264 is the latest digital video codec standard which has proven to be superior than earlier standards in terms of compression ratio quality bit rates and error resilience However unlike Dirac it requires payment of license patent fees The objective of this project is to implement Dirac Pro and H 264 Intra frame coding only video codec encoder and decoder based on several input test sequences and compare there performance Analysis has been done on Dirac Pro and H 264 Intra frame coding only using QCIF and CIF video test sequences as input and the results have been recorded graphically for various parameters including bit rate PSNR SSIM and MSE Acronyms ATSC Advanced television systems committee BBC British Broadcasting Corporation CODEC Encoder decoder DCT Discrete cosine transform GOP Group of picture s HDTV High Definition Television ITU International Telecommunication Union ISO International Standards Organization JPEG Joint Photographic Experts Group JVT Joint Video Team MPEG2 Moving Picture Experts Group MSE Mean square error MV Motion vector QCIF Quarter Common Intermediate Format SSIM Structural similarity index Video Compression Information carrying signals like video may be compressed i e converted to a representation that requires fewer no of bits than the original signal A device or a program that compresses a signal is called an encoder and a device or a program that decompresses the signal is called a decoder An enCOder DECoder pair is called a CODEC The original signal is encoded by the encoder this is called source coding The source coded signal is then encoded further to add error protection this is called channel coding this is done prior to transmitting the signal over the channel At the receiver the channel decoder detects and corrects the transmission errors and source decoder decompresses the signal If the decompressed signal is identical to the original signal its called lossless compression or else its called lossy compression VIDEO ENCODER Video Source Predict Transform Quantize Bitstream as defined by the standard VIDEO DECODER Video Output Reconstruct Inverse transform Encode InverseQuantization Decode Figure 1 Video encoding Decoding processesScope of the standard Dirac is an open and royalty free video codec developed by the BBC named in honor of the British scientist Paul Dirac It aims to provide high quality video compression from web video up to HD 1 Dirac was used internally by the BBC to transmit HDTV pictures at the Beijing Olympics in 2008 2 3 Dirac Pro is a version of the Dirac family of video compression tools optimised for professional production and archiving applications especially where the emphasis is on quality and low latency Dirac Pro is designed for simplicity efficiency and speed and intended for high quality applications with lower compression ratios Like Dirac it is an open technology 4 Dirac Pro will support the following technical features required by professional endusers 4 Intra frame only forward and backward prediction modes are also available if required 10 bit 4 2 2 No subsampling Lossless or Visually lossless compression Low latency on encode decode Robust over multiple passes Ease of transport can use a range of transport standards including MPEG 2 and SDTI Low complexity for decoding Open Specification Multiple vendor Support for multiple HD image formats and frame rates The main difference between Dirac and Dirac PRO is in the treatment of the final process in compression the arithmetic coding Arithmetic coding is processing intensive and introduces delay These are features that are undesirable in high end production work The arithmetic coding produces most efficiency savings with highly compressed material There is little benefit to be gained with the low compression used in top end production DiracPRO therefore omits the arithmetic coding Dirac can compress any size of picture from low resolution QCIF 176x144 pixels to HDTV 1920x1080 and beyond Dirac employs wavelet compression instead of the discrete cosine transforms used in most other codecs such as H 264 MPEG 4 AVC or SMPTE s VC 1 Dirac is one of several projects that have applied wavelets to video compression Wavelet compression has already proven its viability in the JPEG 2000 compression standard for photographic images 5 Dirac Encoder Figure 2 Dirac Encoder Wavelet transform The 2D discrete wavelet transform provides Dirac with the flexibility to operate at a range of resolutions This is because wavelets operate on the entire picture at once rather than focusing on small areas at a time In Dirac the discrete wavelet transform plays the same role of the DCT in MPEG 2 in de correlating data in a roughly frequencysensitive way whilst having the advantage of preserving fine details better This decomposition is repeated to further increase the frequency resolution and the approximation coefficients decomposed with high and low pass filters and then downsampled This is represented as a binary tree with nodes representing a sub space with a different time frequency localisation The tree is known as a filter bank At each level in the figure 3 the signal is decomposed into low and high frequencies Due to the decomposition process the input signal must be a multiple of 2n where n is the number of levels Figure 3 A three level filter bank The choice of filters having compact impulse responses to reduce ringing artifacts caused by wavelets is essential So Daubechies wavelet filters are used to transform and divide the data in sub bands which then are quantized with the corresponding rate distortion optimization RDO parameters and then variable length encoded At the decoder these stages are reversed 6 Figure 4 2 Wavelet transform block diagram 7 The 2 stage decomposition of wavelet transform with only LL sub band is shown in figure 4 Figure 5 3 Stages of wavelet transform 8 Scaling and Quantization Quantization employs a rate
View Full Document
Unlocking...