31.Motion estimation block for HEVC encoder On FPGA,Recent Advances and Innovations

Motion estimation is the most important temporal prediction technique used in video compression. The underlying idea of motion estimation technique can be attributed to the similarity in the consecutive video frames setting apart the changes caused by objects in motion within the frames. It is easy for the encoder to efficiently predict the current frame as a duplicate of the prediction frame in the trivial case of zero motion between frames (and no other differences caused by noise, etc.). Once this is done, the only information needed to transmit to the decoder will be the syntactic overhead necessary for the reconstruction of the picture from the original reference frame.
adaptive offset. Out of these, motion estimation/compensation is the most significant process of video compression as it helps in the removal of the temporal redundancy of the video content to a larger extent. The motion estimation tool finds those motion vectors in the past (or future frames) whose block position matches for every block in the current video frame. These motion vectors are then used in the generation of compensated frames by the motion compensation. It stands out to be the most challenging and time consuming stage in the whole process. Motion estimation of blocks of varying sizes and the presence of multiple reference frames can be counted responsible for this. The cumulative effect of all these factors contributes to the larger encoding time consumed by the ME. This part of the problem could be effectively tackled with the help of efficient algorithms that helps in reducing the overall time consumed.HEVC is designed with the idea of reducing the effective bit-rate down to half of video quality while compared to AVC high profile encoding. The reduction in bit rate could be achieved by employing encoder computation with higher complexity which subsequently requires larger CPU power than the AVC high profile encoding. The rest of the paper is organized as follows. Section II showcases the overview of the motion estimation in HEVC, followed by proposed hardware approach and the corresponding architecture in section III. The implementation results and comparisons are listed in section IV and finally, the conclusion is presented in the section V. II.
Abstract-The world of digital electronics and video compression has become so popular and advanced in such a way that the capture, display and transmission of the ultra-high definition video have become a common thing and is used in various related fields. Due to limited band width and stringent requirements of real-time video play back, video coding is an indispensable process for many visual communication applications and always requires a very high compression ratio. To support this growing popularity and consumer demand, lTU has come up with a new standard HEVC (High Efficiency Video Coding). The new codec helps in achieving the same video quality of the established H.264/A VC codec utilizing just half of the band width, which is notably the greatest advantage it has to offer. Based on the study of motion vector distribution from several commonly used test image sequences, a new diamond search algorithm for fast Motion Estimation (ME) for HECV encoder is proposed in this paper. This algorithm is simple, robust and efficient ensuring low complexity paralleled with high quality in finding global minimum. Keywords- HEVC; Video coding; Motion estimation (ME); Sum of absolute differences (SAD); H.264; Diamond searcll;Coding units(CU).
Mrs.Vidya\ekshmi.V.G Dept. of Electronics and Communication PES Institute of Technology, Bangalore, India vidya0055@yahoo.co.in Prof. Deepa Yagain Dept. of Electronics and Communication PES Institute of Technology, Bangalore, India deepa.yagain@ Mr.Ganesh Rao K Tata Elxsi Ltd. ITPB Main Road, Bangalore, India.
MOTION ESTIMATION IN HEVC
I.
INTRODUCTION
Motion estimation eliminates temporal redundancy information between successive frames there by enhancing the encoding efficiency. This has been adopted by various video­ coding standards. HEVC is a state-of-the art standard in video compression rendering more accurate prediction and higher compression efficiency. In the years to come, this new standard HEVC [1] would be impacting on a large scale the areas of mobile communication, television broadcasting, multimedia streaming and video capture & storage areas. HEVC standard is merited with features like lower memory bandwidth, less average interpolation complexity (due to the 1D directional interpolation filters) and twice the compression efficiency in comparison with its predecessors H.264/MPEG-4 AVC[2]. HEVC has been proposed with a vision to minimize the complexity and computation load involved in the motion estimation. The various features of HEVC are coding units and coding blocks, prediction units and prediction blocks, transform units and transform blocks, motion vector signaling, motion compensation, intra-picture prediction, quantization control, entropy coding, in-loop deblocking filtering and sample