Background: Advances in video compression technology have been driven by ever-increasing processing power available in software and hardware. Methods: The emerging High-Efficiency Video Coding (HEVC) standard aims to ...
详细信息
Background: Advances in video compression technology have been driven by ever-increasing processing power available in software and hardware. Methods: The emerging High-Efficiency Video Coding (HEVC) standard aims to provide a doubling in coding efficiency with respect to the H.264/AVC high profile, delivering the same video quality at half the bit rate. Results: Thus, the results show high computational complexity. In both standards, the motion estimation block presents a significant challenge in clock latency since it consumes more than 40% of the total encoding time. For these reasons, we proposed an optimized implementation of this algorithm on a low-cost NVIDIA GPU developed with CUDA language. Conclusion: This optimized implementation can provide high-performance video encoder where the speed reaches about 85.
In order to speed up H.264/AVC coding efficiency, this paper proposed a parallelization approach of fullsearch (FS) algorithm for motion estimation on Graphic Processor Unit (GPU) using computing unified device archi...
详细信息
Motion estimation is the computationally intensive part of video encoding. This paper presents a processing element based architecture for accelerating the calculation of Sum of Absolute Differences (SAD) which is the...
详细信息
Motion estimation is the computationally intensive part of video encoding. This paper presents a processing element based architecture for accelerating the calculation of Sum of Absolute Differences (SAD) which is the most widely used block matching criteria in motion estimation. A clock gating method is anlysed to enable or disable the required processing elements for a particular time of use. The selection of processing elements is performed based on motion analysis of the input video. The level of motion is estimated from initial frames to configure the hardware for SAD evaluation. A System-on-Chip approach, implemented in Xilinx Zynq SoC is proposed that will be efficient in terms of power and resource utilization as the hardware is configured based on the property of input video. This hardware-software co-design is able to achieve approximately 4.6x speed up compared to the the original software implementation of the framework running on ARM processor. (C) 2020 Elsevier B.V. All rights reserved.
According to the observation on the distribution of motion differentials among the motion vector of any block and those of its four neighboring blocks from six real video sequences, this paper presents a new predictiv...
详细信息
According to the observation on the distribution of motion differentials among the motion vector of any block and those of its four neighboring blocks from six real video sequences, this paper presents a new predictive search area approach for fast block motion estimation. Employing our proposed simple predictive search area approach into the fullsearch (FS) algorithm, our improved FS algorithm leads to 93.83% average execution-time improvement ratio, but only has a small estimation accuracy degradation. We also investigate the advantages of computation and estimation accuracy of our improved FS algorithm when compared to the edge-based searchalgorithm currently published by Chan and Sin;experimental results reveal that our improved FS algorithm has 74.33% average execution-time improvement ratio and has a higher estimation accuracy. Finally, we further compare the performance among our improved FS algorithm, the three-step searchalgorithm, and the block-based gradient descent searchalgorithm.
This study proposes two efficient block matching algorithms for fast motion estimation using a histogram ordering model in order to reduce the computational cost of video coding. Since the representative pixels that c...
详细信息
This study proposes two efficient block matching algorithms for fast motion estimation using a histogram ordering model in order to reduce the computational cost of video coding. Since the representative pixels that consist of edge or texture mainly contribute to the block matching criterion, it is important to analyse the details of the current block. To analyse the characteristics of the block with low complexity, the pixel histogram is used for the observation of the block. Based on this pixel histogram for the current block, an optimal block matching order is determined for the lossless fast matching algorithm. In addition, adaptive partial block matching algorithm for lossy fast motion estimation using histogram-based block matching ordering is also presented to further reduce the complexity of block matching. By capturing the representative pixels, block matching distortion is maximised soon enough during the block matching. For this reason, partial block matching is available instead of full block matching, which is also known as a full search algorithm. Our experimental results show that the proposed algorithm not only reduces the computational complexity of block matching criterion, but also tends to maintain the image quality when compared to the conventional fast matching algorithms.
Block-based motion estimation is essential in video coding applications, which can diminish the temporal redundancy of sequences effectively. Though full search algorithm (FSA) can undoubtedly get global optimum estim...
详细信息
ISBN:
(纸本)0780377028
Block-based motion estimation is essential in video coding applications, which can diminish the temporal redundancy of sequences effectively. Though full search algorithm (FSA) can undoubtedly get global optimum estimation, it is computationally intensive. In this paper a novel algorithm named as the Sampling Predictive Block Match algorithm (SPBMA) is presented, which combine the ideas of motion vector mean predictor sampling match and adaptive diamond search. Our experiments verify the superiority of the proposed algorithm in precision and rapidity (for the Mother & daughter sequence, SPBMA is 108 times as fast as FSA with only 0.75% lower in PSNR). It is suitable for the wireless video communication with the rigorous demand for ultra low power.
Based on a study of the motion vector (MV) obtained by the full search algorithm among several test sequences, we found that the prediction process for motion estimation in many existing algorithms are not optimal. We...
详细信息
ISBN:
(纸本)0780374029
Based on a study of the motion vector (MV) obtained by the full search algorithm among several test sequences, we found that the prediction process for motion estimation in many existing algorithms are not optimal. We optimize the prediction process by classifying the blocks into four different classes based on a newly proposed three-dimensional reference MV. Different search patterns and strategies are used for each class. The proposed algorithm can improve computation requirement as high as 43.9% than the diamond searchalgorithm.
The paper presents a novel Orthogonal Logarithmic search (OLS) method for block base motion compensation. The performance of the algorithm is evaluated, by using standard QCIF benchmark video sequences and the results...
详细信息
ISBN:
(纸本)0819448222
The paper presents a novel Orthogonal Logarithmic search (OLS) method for block base motion compensation. The performance of the algorithm is evaluated, by using standard QCIF benchmark video sequences and the results are compared to a traditional well-known full search algorithm (FSA) and a sub-optimal method called the Three Step search (3SS). The evaluation considers the three important metrics, time, entropy and PSNR (Peak Signal to Noise Ratio). The paper also shows that the strength of the algorithm lies in its speed of operation as it is 95% faster than the FSA and over 60% faster than the 3SS. Investigation into-motion estimation algorithms is one of the important issues in the video coding standards such as ISO MPEG-1/2 and ITU-T H.263. These international standard's regularly use a conventional FSA to estimate the motion of pixels between pairs of image block matching algorithms. Since a fullsearch requires intensive computations and the distortion function needs to be evaluated many times for each target block to be matched, the process is very time consuming. Therefore, the main aim of this investigation has been to alleviate this acute problem of search speed and accuracy.
This paper presents a fast efficient codeword searchalgorithm for vector quantization (VQ) based on the Adaptive Multi-Rate Wideband (AMR-WB) speech codec. AMR-WB is the speech codec standardized by the 3GPP for the ...
详细信息
ISBN:
(纸本)9781424455690
This paper presents a fast efficient codeword searchalgorithm for vector quantization (VQ) based on the Adaptive Multi-Rate Wideband (AMR-WB) speech codec. AMR-WB is the speech codec standardized by the 3GPP for the GSM and WCDMA 3G system. It is also selected as the ITU-T G. 722.2 standards. In AMR-WB speech codec VQ codeword search section, full search algorithm is used to find matching codeword. This requires a huge amount of computation. For large-size codebooks and high-dimensional vector of AMR-WB, computational complexity is significantly higher. The proposed fast codeword search method selects the codeword, which has the least absolute error component in VQ codebook, as the initial matching codeword. It can help to reduce computational complexity, whereas the reconstructed speech quality almost has no dropping. The article introduces the basic principle of the fast codeword searchalgorithm in detail. Simulation results show that the algorithm based on AMR-WB speech codec, can reduce the time of vector quantization codebook search and the w-PESQ of nine modes is no significant dropping.
Based on motion estimation using spatio-temporal correlations, an enhanced searchalgorithm is proposed to further improve the performance in terms of reducing total amount of computations while preserving the video q...
详细信息
ISBN:
(纸本)0819456586
Based on motion estimation using spatio-temporal correlations, an enhanced searchalgorithm is proposed to further improve the performance in terms of reducing total amount of computations while preserving the video quality. Our experimental results show that the proposed algorithm has substantially smaller computational complexity having some quality reduction. For sequences with low motion PSNR reduction is around 0.1 dB and computational complexity is around 1700 times less than that of a fullsearch.
暂无评论