Transcoding from H.264/AVC to High Efficiency Video Coding (HEVC)

EE5359 PROJECT INTERIM REPORT Transcoding from H.264/AVC to High Efficiency Video Coding (HEVC) Shantanu Kulkarni UTA ID: 1000789943

Transcoding from H.264/AVC to HEVC Objective: To discuss and implement HEVC transcoder on AVC bitstream to obtain HEVC standardized compression [1] Introduction The strategies to employ the transcoding methods from H.264/AVC [2] bit stream to HEVC[12] are discussed in this paper. HEVC and AVC share similar coding architecture. For inter picture, the power spectrum based rate distortion optimization model as well as input residual, modes and motion vectors are utilized to estimate the best coding unit split code-tree, the best prediction unit (PU) mode and the best motion vector estimation for each PU. Also for intra frame prediction, it is required to reduce the CU and PU partition candidate. Design of most video coding standards is primarily aimed at having the highest coding efficiency, which is the ability to encode the video at lowest possible bitrate while maintaining certain level of video quality. HEVC which is a recently emerged video coding standard aims at high coding efficiency while retaining the video quality. With its hybrid coding architecture, motion compensation prediction and transform coding technique, it can be seen as an improved version of the previous standard H.264. Transcoding from H.264 to HEVC will enable lowering the bitrate resulting in a more efficient compression. The proposed transcoding technique can achieve a good tradeoff between good tradeoff between coding efficiency and trancoding complexity. The increasing of networked video applications, e.g. video conferencing, IPTV and HDTV, with resolution ranging from QVGA to ultra-high definition video, have posed new challenges to design video representation and transmission system, especially for applications with various devices through heterogeneous wired and wireless networks. How to make video be suitable for various device capabilities and dynamical bandwidths becomes very challenging. Transcoding is one of the most promising technologies, which provides video adaptation in terms of bit-rate reduction, resolution reduction and format conversion to meet various requirements. The HEVC standard is based on the well-known block-based hybrid coding architecture, combining motion-compensated prediction and transform coding with high-efficiency entropy coding. However, in contrast to previous video coding standards, it employs a flexible quad-tree coding block partitioning structure that enables the efficient use of large and multiple sizes of coding, prediction, and transform blocks. It also employs improved intra prediction and coding, adaptive motion parameter prediction and coding, new loop filter and an enhanced version of context-adaptive binary arithmetic coding (CABAC) entropy coding. New high level structures for parallel processing are also employed. 2

The ITU-T began development of a successor to H.264 in 2004, while ISO/IEC (Chapter 5: HEVC) began working in 2007. In January 2010, the groups collaborated on a joint Call for Proposals, which culminated in a meeting of the MPEG & VCEG Joint Collaborative Team on Video Coding (JCT-VC) in April 2010, when the name High Efficiency Video Coding (HEVC) was adopted for the codec. [2] Transcoder Structure and Basics Video transcoding is the operation of converting video from one format to another. A format is defined by characteristics such as bit-rate, spatial resolution etc. One of the earliest applications of transcoding is to adapt the bit-rate of a compressed stream to the channel bandwidth for universal multimedia access in all kinds of channels like wireless networks, internet, dial-up networks etc. Changes in the characteristics of an encoded stream like bit-rate, spatial resolution, quality etc can also be achieved by scalable video coding. However, in cases where the available network bandwidth is insufficient or if it fluctuates with time, it may be difficult to set the base layer bit-rate. In addition, scalable video coding demands additional complexities at both the encoder and the decoder. The emerging developments in video coding technology make transcoding much more complicated. A new developed video coding standard always creates new requirement for transcoding from existed formats to the new format for the interoperability of video contents. Figure 2 is the most basic transcoding architecture. The motion vectors from the incoming bit stream are extracted and reused. Thus the complexity of the motion estimation block is eliminated which accounts for 60% of the encoder computation. Hence, even though it is slightly more complex, it is suited for heterogeneous transcoding between different standards where the basic parameters like mode decisions, motion vectors etc are to be re-derived. 3

[3] Overview of HEVC Fig. 1 Encoder block diagram HEVC [12] Fig. 2 Transcoded pixel domain transcoding architecture [4] The input video is first divided into blocks called coding tree units (CTUs), which perform a role that is broadly analogous to that of macroblocks in previous standards. The coding unit (CU) defines a region sharing the same prediction mode (intra, inter or skip) and it is represented by the leaf node of a quadtree structure. The prediction unit (PU) defines a region sharing the same prediction information. The transform unit (TU), specified by another quadtree, defines a region sharing the same transformation and quantization. 4

Fig. 3 Recursive block structure for HEVC [13] Fig. 4 Intra prediction mode, HEVC [12] The best intra mode among a total of 35 modes (Fig. 4) (Planar, DC and 33 angular directions) is selected and coded. Mode dependent context sample smoothing is applied to increase prediction efficiency and the three most probable modes (MPM) are used to increase symbol coding efficiency. The best motion parameters are selected and coded by merge mode and adaptive motion vector prediction (AMVP) mode, in which motion predictors are selected and explicitly coded among several candidates. To increase the efficiency of motion-compensated prediction, non-cascaded interpolation structure with 1D FIR filters are used. An 8-tap or 7-tap filter is directly applied to generate the samples of half-pel and quarter-pel luma samples, respectively. A 4-tap filter is utilized for chroma interpolation. Residuals generated by subtracting the prediction from the input are spatially transformed and quantized. In the transform process, matrices which are approximations to DCT are used. For low computational cost, partial butterfly structure is implemented for transformation. In the case of 4x4 intra predicted 5

residuals, DST is used for luma. 52-level quantization steps and rate-distortion optimized quantization (RDOQ) are used in the quantization process. Reconstructed samples are created by inverse quantization and inverse transform. CABAC encoding scheme is used in this encoding standard, which is applied to the generated symbols and quantized transform coefficients. After reconstruction, two in-loop filtering processes are applied to achieve better coding efficiency and visual quality: deblocking filtering and sample adaptive offset (SAO). Reconstructed CTUs are assembled to construct a picture and stored in the decoded picture buffer to be used to encode the next picture of input video. [4] Overview of H.264/AVC H.264 [2] is a standard for video compression, and is equivalent to MPEG-4 Part 10, or MPEG-4 AVC (for advanced video coding). As of 2008, it was the latest block-oriented motion-compensation-based video standard developed by the ITU-T Video coding experts group (VCEG) together with the ISO/IEC moving picture experts group (MPEG), and it was the product of a partnership effort known as the joint video team (JVT). The ITU-T H.264 standard and the ISO/IEC MPEG-4 part 10 standard (formally, ISO/IEC 14496-10) are jointly maintained so that they have identical technical content. Fig. 5 H.264 encoder [2] Fig. 6 H.264 decoder [2] 6

Features for enhancement of prediction are as follows. Directional spatial prediction for intra coding (9 directional prediction modes) Variable block-size motion compensation with small block size Quarter-sample-accurate motion compensation Motion vectors over picture boundaries Multiple reference picture motion compensation Decoupling of referencing order from display order Decoupling of picture representation methods from picture referencing capability Weighted prediction Improved skipped and direct motion inference In-the-loop deblocking filtering Features for improved coding efficiency are as follows. Small block-size transform Exact-match inverse transform Short word-length transform Hierarchical block transform Arithmetic entropy coding Context-adaptive entropy coding Features for robustness to data errors/losses are as follows. Parameter set structure NAL unit syntax structure Flexible slice size Flexible macroblock ordering (FMO) Arbitrary slice ordering (ASO) Redundant slices (RS) Data partitioning SP/SI synchronization/switching pictures [5] Comparison with H.264 and previous standards Owing to a number of diverse applications/fields which have been introduced in HEVC, it may overtake the previous coding standard H.264/AVC. Following are the areas in which the AVC and HEVC have differences in their fields: Larger block structure leading to maximum of 64x64 pixels per block Intra prediction direction modes which are upto 35 (33 modes + DC + Planar) in case of HEVC while H.264 has 9 directional modes of intra prediction (Fig. 7) Adaptive motion vector prediction, which allows codec to find more inter frame redundancies 7

Superior parallelization tools, including wavefront parallel processing, for more efficient coding in a multi core environment Entropy using CABAC only, no more CAVLC Improvements to de-blocking filter and addition of one more filter called Sample Adaptive Offset (SAO) that further leaves artifacts along block edges Reduction of bit rate by almost 37% (Approximate) Fig. 7: Intra predicton modes, H.264/AVC [20] The differences between AVC and HEVC can be summarized through the following table: Table1: Difference between H.264/AVC and HEVC [1] 8

[6] HEVC Transcoding This topic discusses several transcoding strategies for AVC to HEVC transcoding with bit rate reduction. Considering the similar coding architecture of HEVC and AVC, and motivated by the work in [1], for inter picture transcoding, PS-RDO model is utilized to determine the CU quadtree structure, the best CU partition mode and the best motion vector of each prediction unit (PU), and for intra picture transcoding, it is proposed to reduce the candidate settings for CU quadtree structures and PU partitions. The transcoding schemes discussed here avoid high computational complexity in terms of reduced RDO evaluations and motion compensation operation as well as fractional pixel interpolation operation. The pixel domain AVC to HEVC architecture is illustrated in Fig. 8 Input AVC Bitstream AVC Decoder HEVC Re-encoder Output HEVC Bitstream Residual, modes and MVs CU, PU partitions and MVs Simplified Mode Selection Fig. 8 Pixel domain AVC-HEVC transcoder [1] Transcoding of intra Coded Frames The quality of each intra picture will have significant impacts on the inter coded frames. Thus, its quality needs to be kept as intact as possible. The input AVC bitstream already contains useful information of the MB partitions and prediction directions, the LCU will be initially split according to input macroblocks to the AVC. The CU partitions needs to be further merged into larger sizes according to the predicted directions of the neighboring PUs. The encoding computational complexity in HEVC to select the best coding parameter has been increased as the increasing of candidates for CU partitions, PU partitions and TU partitions. The intra predictor and motion vector predictor, of a CU is generated from neighboring coded CUs, the HM 4 performs preorder recursive traversal on the CU quadtree. When encoding a CU at current depth, the best PU mode is determined by successively evaluating the RD costs of each inter and intra modes. Indicates that this mode be evaluated conditionally, either depends on the CU size or some fast mode decision algorithms. After the best PU mode is determined, if current CU is larger than 8x8, it might further split into four sub-cus, and then recursively calls the CU compressing function to determine the best CU quadtree structure. The 9

decision of best TU split tree is integrated in the determination of the best PU mode. Fig. 9 CU compressing in HEVC encoder [1] Transcoding of Inter Coded Frames The major complexity of Inter picture coding comes from the motion estimation (ME), MC, T/Q and IQ/IT operations when testing every set of possible coding parameters with possible CU size, PU and TU modes. Thus, these operations can be reduced by utilizing the information directly from the AVC encoded format. The information that can be used are motion vectors to decide the displacement, the residuals and the modes of the predictions. The key technology of AVC to HEVC inter picture transcoding is to merge smaller blocks to a larger CU, especially for bit rate reduction transcoding. Since a large CU may consists of different 4x4 blocks, and probably, these blocks may have different MVs, merging these blocks now turns to measure the RD cost when the MV changes. [7] Video transcoding architectures Cascaded Architecture- As the first step towards development of the transcoder, the cascaded decoderencoder architecture is used. The H.264 coded bitstream is decoded using H.264 decoder and the reconstructed video sequences are encoded using HEVC encoder. The output of the transcoder is the HEVC bitstream. 10

Cascaded Decoder and Encoder Transcoder Input Video Sequence H.264 Encoder H.264 bitstream H.264 Decoder Reconstructed sequence HEVC Encoder Transcoded HEVC Bit stream HEVC Decoder Output reconstructed video sequence Fig. 10 Cascaded transcoder architecture (Cascaded Decoder-Encoder) This type of implementation involves complete decoding and re-encoding of the incoming compressed video stream. It has to perform full decoding followed by the resizing / reordering of the decoded sequence before re-encoding it [4]. Due to complete re-encoding operation, complex frame reordering and full-scale motion re-estimation are required. Motion estimation has the highest complexity in the encoder. So such an implementation involves the maximum complexity and also high processing time and power consumption leading to significant delays. Also the pictures / frames exhibit increased error due to re-encoding being performed on decoded pictures which have lower quality than original frames. The error is due to propagation. Lossy encoding process inserts errors; when such a bitstream is decoded, the decoded pictures have errors which propagate on further encoding which inserts more errors. This error is not the same as the drift error in open loop transcoders. Due to all these reasons such a transcoding model needs a lot of optimization. [17] The four video sequences are used for transcoding from H.264 to HEVC format: Akiyo_qcif, Foreman_qcif, mobile_cif and coastguard_cif. All of these are standard YUV sequences in 4:2:0 formats. The cif sequences have resolution 352 x 288 pixels while qcif sequences have resolution 176 x 144 pixels. The simulations have been carried out at a fixed bitrate by keeping the quantization parameter constant. Table 2 displays the MSE and PSNR values for akiyo_qcif video sequence while it is transcoded using cascaded decoder-encoder transcoder architecture using 100 frames of the sequence. Similarly, Table 3, Table 4 and Table 5 display the PSNR and the MSE values of the video frames foreman_qcif, mobile_cif and coastguard_cif respectively. The PSNR for the colored images for sequence of 100 frames is calculated as: PSNR for colored images = 1/8 * (6*Ypsnr + Upsnr + Vpsnr ) Also, the computation time and the bitates obtained for the respective video sequences are displayed in Tables 2, 3, 4 and 5. 11

Sequence Component Metric akiyo_qcif Y U V Encoded by H.264 Encoded by HEVC Transcoded output with respect to original Transcoded output with respect to H.264 MSE 7.9453 16.3527 14.9827 13.0089 PSNR 39.14 35.9949 36.3749 36.9884 MSE 4.89645 73.6801 7.6814 5.4491 PSNR 41.234 39.4573 39.2764 40.7675 MSE 4.05427 44.4938 5.099 3.8215 PSNR 42.054 41.6478 41.056 42.3084 PSNR color 39.766 37.13431 37.322725 38.1257875 Bitrate (kbps) 15.53 12.4128 11.64 Computation Time (sec) 149.847 504.537 494.878 Table 2. MSE and PSNR on akiyo_qcif.yuv video sequence 100 frames Sequence Component Metric foreman_qcif Y U V Encoded by H.264 Encoded by HEVC Transcoded output with respect to original Transcoded output with respect to H.264 MSE 16.03446 29.9875 16.0345 23.6761 PSNR 36.134 33.3614 36.0803 34.3877 MSE 5.56395 7.3461 5.5639 3.5923 PSNR 40.689 39.4702 40.677 42.577 MSE 4.00861 5.8822 4.0086 3.9526 PSNR 42.124 40.4354 42.1009 42.1619 PSNR color 37.452125 35.00925 37.4074625 36.3831375 Bitrate (kbps) 90.6 45.8904 43.4808 Computation Time (sec) 198.281 851.182 839.835 Table 3. MSE and PSNR on foreman_qcif.yuv video sequence 100 frames Sequence Component Metric mobile_cif Y U V Encoded by H.264 12 Encoded by HEVC Transcoded output with respect to original Transcoded output with respect to H.264 MSE 33.74791 60.16589 29.5597 48.6284 PSNR 33.021 30.3373 33.4238 31.2619 MSE 16.33582 19.1911 24.0076 9.5459 PSNR 36.033 35.2998 34.3273 38.3326 MSE 17.23845 22.2633 26.3195 12.7897 PSNR 35.806 34.6549 33.928 37.0622 PSNR color 33.745625 31.49731 33.5997625 32.870775 Bitrate (kbps) 851.1 361.4736 337.5528 Computation Time (sec) 605.525 4053.218 3957.333 Table 4. MSE and PSNR on mobile_cif.yuv video sequence 100 frames

Sequence Component Metric coastguard_cif Y U V Encoded by H.264 Encoded by HEVC Transcoded output with respect to original Transcoded output with respect to H.264 MSE 43.65598 54.3388 33.7838 32.0597 PSNR 31.797 30.7797 32.8437 33.0712 MSE 3.745 3.7246 21.7963 1.75509 PSNR 42.499 42.42 34.747 45.6878 MSE 3.11601 2.9498 24.5536 1.4232 PSNR 43.26 43.4328 34.2296 46.5979 PSNR color 34.567625 33.81638 33.25485 36.3391125 Bitrate (kbps) 428.1 295.2936 233.5488 Computation Time (sec) 846.241 4012.575 3791.309 Table 5. MSE and PSNR on Coastguard_cif.yuv video sequence 100 frames Fig 11, Fig 12, Fig 13, Fig 14 display the bitrate comparison between the H.264 encoded output, HEVC encoded output and cascaded transcoder output using akiyo_qcif, foreman_qcif, mobile_cif and coastguard_cif video sequences respectively. For each video sequence, 100 frames were taken into consideration. Bit Rate (kbps) 16 15.5 15 14.5 14 13.5 13 12.5 12 11.5 11 10.5 10 Bitrate_akiyo_qcif Video Sequence Encoded by H.264 Encoded by HEVC Transcoded output with respect to H.264 reconstructed frames Fig. 11 Bitrate comparison between H.264 encoded, HEVC encoded and transcoded output using akiyo_qcif.yuv sequence. (100 frames) 13

Bit Rate (kbps) 100 95 90 85 80 75 70 65 60 55 50 45 40 Bitrate_foreman_qcif Video Sequence Encoded by H.264 Encoded by HEVC Transcoded output with respect to H.264 reconstructed frames Fig. 12 Bitrate comparison between H.264 encoded, HEVC encoded and transcoded output using foreman_qcif.yuv sequence. (100 frames) Bit Rate (kbps) 900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 Bitrate_mobile_cif Video Sequence Encoded by H.264 Encoded by HEVC Transcoded output with respect to H.264 reconstructed frames Fig. 13 Bitrate comparison between H.264 encoded, HEVC encoded and transcoded output using mobile_cif.yuv sequence. (100 frames) 14

Bit Rate (kbps) 450 425 400 375 350 325 300 275 250 Encoded by H.264 Encoded by HEVC Transcoded output with respect to H.264 reconstructed frames 225 200 Video Bitrate_coastguard_cif Sequence Fig. 14 Bitrate comparison between H.264 encoded, HEVC encoded and transcoded output using coastguard_cif.yuv sequence. (100 frames) Fig 15 and Fig. 16 display graphically the comparison between the PSNR for colored images values between the H.264 encoded output, HEVC encoded output and cascaded transcoded output with respect to original sequence and cascaded transcoder output with respect to H.264 reconstructed sequences, using akiyo_qcif and foreman_qcif (Fig 15 ) and mobile_cif and coastguard_cif (Fig. 16) video sequences. For each video sequence, 100 frames were taken into consideration. PSNR (db) 40 39.5 39 38.5 38 37.5 37 36.5 36 35.5 35 34.5 34 Encoded by H.264 Encoded by HEVC Transcoded output with respect to original Transcoded output with respect to H.264 reconstructed frames Video PSNR_akiyo_qcif PSNR_foreman_qcifSequences Fig. 15 PSNR comparison between H.264 encoded, HEVC encoded, transcoded output with respect to original sequences and transcoded output with respect to H.264 reconstructed sequences using akiyo_qcif.yuv and foreman_qcif.yuv sequences. (100 frames) 15

PSNR (db) 37 36.5 36 35.5 35 34.5 34 33.5 33 32.5 32 31.5 31 30.5 30 PSNR_mobile_cif PSNR_coastguard_cif Video Sequences Encoded by H.264 Encoded by HEVC Transcoded output with respect to original Transcoded output with respect to H.264 reconstructed frames Fig. 16 PSNR comparison between H.264 encoded, HEVC encoded, Transcoded output with respect to original sequences and transcoded output with respect to H.264 reconstructed sequences using mobile_cif.yuv and coastguard_cif.yuv sequences. (100 frames) Time (sec) 550 525 500 475 450 425 400 375 350 325 300 275 250 225 200 175 150 125 100 Computation Time Akiyo Video Sequences Encoded by H.264 Encoded by HEVC Transcoded output with respect to H.264 reconstructed frames Fig. 17 Comparison of computation time between H.264 encoded, HEVC encoded and transcoded using akiyo_qcif.yuv sequence. (100 frames) 16

Time (sec) 900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 Computation Time Foreman Video Sequences Encoded by H.264 Encoded by HEVC Transcoded output with respect to H.264 reconstructed frames Fig. 18 Comparison of computation time between H.264 encoded, HEVC encoded and transcoded using foreman_qcif.yuv sequence. (100 frames) Time (sec) 4000 3750 3500 3250 3000 2750 2500 2250 2000 1750 1500 1250 1000 750 500 Computation Time Mobile Video Sequences Encoded by H.264 Encoded by HEVC Transcoded output with respect to H.264 reconstructed frames Fig. 19 Comparison of computation time between H.264 encoded, HEVC encoded and transcoded using mobile_cif.yuv sequence. (100 frames) 17

Time (sec) 4250 4000 3750 3500 3250 3000 2750 2500 2250 2000 1750 1500 1250 1000 750 500 Computation Time Coastguard Video Sequences Encoded by H.264 Encoded by HEVC Transcoded output with respect to H.264 reconstructed frames Fig. 20 Comparison of computation time between H.264 encoded, HEVC encoded and transcoded using Coastguard_cif.yuv sequence. (100 frames) Fig. 21 Akiyo_qcif H.264 decoder output Fig. 22 Akiyo_qcif transcoder output Fig. 23 Akiyo_qcif HEVC decoder output 18

Fig. 24 Foreman_qcif H.264 decoder output Fig. 25 Foreman_qcif transcoded output Fig. 26 Foreman_qcif HEVC encoded and reconstructed Fig. 27 Mobile_cif H.264 encoded and reconstructed 19

Fig. 28 Mobile_cif Transcoded output Fig. 29 Mobile_cif HEVC Encoded and reconstructed 20

Fig. 30 Coastguard_cif H.264 encoded and reconstructed Fig. 31 Coastguard_cif Transcoded Output 21

Fig. 32 Coastguard_cif HEVC encoded and reconstructed From Fig11, Fig12, Fig13, Fig14, it is seen that the reduction in the bitrate obtained after transcoding is minimal. Which means that the reduction might be due to the changes in content of the sequences. After reconstruction of the video by decoding the H.264 bit streams, some of the content is lost, which reflects in the HEVC encoded bitrate. The computation time graph shows a slight improvement in the computation time after transcoding. The PSNR values from Fig 15 and Fig 16 shows that the trasncoded output is slightly better than the one decoded by the HEVC encoder alone. Also, there is a slight improvement in the computation time of the encoder, which can be seen in the Fig17, Fig 18, Fig 19 and Fig. 20. [8] Next Steps in the Project Next efforts in this project would be to import the motion vectors from the H.264 decoder and use them in HEVC encoder. The maximum computation time is required by the motion vector estimation. If it is possible to import the motion vectors directly from the H.264, then it would reduce the time required for computation along with some improvement in bitrate. Input AVC Bitstream AVC Decoder HEVC Re-encoder Output HEVC Bitstream Reuse the MV and MVs Save the motion vectors Fig 33 Transcoder Architecture [4] 22

[9] Conclusion: Transcoding strategies for AVC to HEVC transcoding with bitrate reduction are proposed in this paper. With the input residual, modes and motion vectors of AVC, the PS-RDO model is utilized to determine the best coding unit splitting quadtree, the best prediction unit and the best motion vector. The number of required RDO evaluations is significantly reduced for both intra and inter picture transcoding. Besides, the motion estimation, motion compensation as well as fractional pixel interpolation operations are avoided in the proposed inter picture transcoding strategy. The proposed transcoding strategies maintain good tradeoff between coding efficiency and transcoding complexity. References: [1] D. Zhang, B. Li, J. Xu, and H. Li, Fast Transcoding from H.264/AVC to High Efficiency Video Coding IEEE International Conference on Multimedia Expo, pp. 651-656, July, 2012 [2] T. Wiegand et al, Overview of the H.264/AVC video coding standard, IEEE Trans. CSVT, Vol. 13, pp. 560-576, July 2003. [3] J Xin, C.W. Lin and M.T. Sun, Digital video transcoding, Proceedings of the IEEE, Vol. 93, pp 84-97, Jan 2005. [4] A. Vetros, C. Christopoulos and H. Sun, Video transcoding architectures and techniques: An overview, IEEE Signal Processing Magazine, Vol. 20, pp 18-29, March 2003. [5] S. Matsuo, S. Takamura and A. Shimizu, Modification of Intra Angular Prediction in HEVC IEEE, Signal & Information Processing Association Annual Summit and Conference (APSIPA ASC), 2012 Asia-Pacific, pp 1-4, Dec 2012. [6] D. Zhang, B. Li, J. Xu, and H. Li, Fast Transcoding from H.264/AVC to High Efficiency Video Coding IEEE International Conference on Multimedia Expo, pp. 651-656, July, 2012 [7] T. Wiegand, G. J. Sullivan, G. Bjøntegaard, and A. Luthra, Overview of the H.264/AVC Video Coding Standard, IEEE transactions on circuits and systems for video technology, vol. 13, no. 7, pp. 560-576, July 2003 [8] I. Kim, J. Min, T. Lee et al, Block Partitioning Structure in the HEVC Standard, IEEE transactions on circuits and systems for video technology, vol. 22, no. 12, pp. 1697-1706, December 2012 [9] Q. Cai, L. Song, G. Li et al, Lossy and Lossless Intra Coding Performance Evaluation: HEVC, H.264/AVC, JPEG 2000 and JPEG LS, Asia Pacific Signal & Information Processing Association Annual Summit and Conference (APSIPA ASC), pp. 1-9, Dec 2012. 23

[10] G. Sullivan, P. Topiwalla and A. Luthra, The H.264/AVC video coding standard: overview and introduction to the fidelity range extensions, SPIE Conference on Applications of Digital Image Processing XXVII, vol. 5558, pp. 53-74 Aug 2004. [11] T. Weigand et al, Introduction to the Special Issue on Scalable Video Coding Standardization and Beyond IEEE Trans on Circuits and Systems for Video Technology, Vol 17, pp 1099-1102, Sept 2007. [12] T. D. Nguyen et al, Efficient MPEG-4 to H.264/AVC transcoding with spatial downscaling, ETRI Journal, vol.29, no.6, pp 826-828, Dec. 2007. [13] G.J. Sullivan, J. Ohm, W. Han et al, Overview of High Efficiency Video Coding (HEVC) Standard IEEE Transactions on Circuits and Systems for Video Technology, Vol. 22, No.12, Dec 2012 [14] H. Zhang and Z. Ma, Fast intra prediction for high efficiency video coding, Pacific Rim Conf. on Multimedia, PCM2012, Singapore, Dec. 2012. [15] HEVC open source software (encoder/decoder) https://hevc.hhi.fraunhofer.de/svn/svn_hevcsoftware/tags/hm-6.0 [16] JM Reference Software - http://iphome.hhi.de/suehring/tml/ [17] Eduardo Peixoto Fernandes da Silva, Advanced Heterogeneous Video Transcoding Queen Mary, University of London, PhD Thesis. [18] J. Padia, Complexity Reduction For Vp6 To H.264 Transcoder Using Motion Vector Reuse, MPL, University of Texas at Arlington, May 2012. Reference Books [19] K. Sayood, Introduction to Data compression, III edition, Morgan Kaufmann publishers, 2006. [20] I.E.G. Richardson, H.264 and MPEG-4 video compression: video coding for next-generation multimedia, Second Edition, Wiley, 2010 Websites [21] http://en.wikipedia.org/wiki/ : Website for Wikipedia, Encyclopedia [22] http://www-ee.uta.edu/dip/courses/ee5359/index.html: Course website [23] http://ieeexplore.ieee.org/: Website archive for IEEE papers online [24] http://www.v-net.tv/hevc-is-game-changer-for-multi-screen-and-iptv/: Impact of HEVC standard on digital media market like cell phones, TVs etc [25] http://www.streamingmedia.com/articles/editorial/what-is-.../what-is- HEVC-(H.265)-87765.aspx: Summary about HEVC, information site. 24

[26] http://mrutyunjayahiremath.blogspot.com/2010/09/h264-videocodec_22.html: Diagram for H.264 prediction direction modes [27] http://codesequoia.wordpress.com/2012/10/28/hevc-ctu-cu-ctb-cb-pb-andtb/ : Block coding in HEVC. Also link to make a HEVC stream. 25