• ETRI Journal
• /
• 제31권6호
• /
• pp.784-794
• /
• 2009

We present a full HD (1080p) H.264/AVC High Profile hardware encoder based on fast motion estimation (ME). Most processing cycles are occupied with ME and use external memory access to fetch samples, which degrades the performance of the encoder. A novel approach to fast ME which uses shared multibank memory can solve these problems. The proposed pixel subsampling ME algorithm is suitable for fast motion vector searches for high-quality resolution images. The proposed algorithm achieves an 87.5% reduction of computational complexity compared with the full search algorithm in the JM reference software, while sustaining the video quality without any conspicuous PSNR loss. The usage amount of shared multibank memory between the coarse ME and fine ME blocks is 93.6%, which saves external memory access cycles and speeds up ME. It is feasible to perform the algorithm at a 270 MHz clock speed for 30 frame/s real-time full HD encoding. Its total gate count is 872k, and internal SRAM size is 41.8 kB.

• Biotechnology and Bioprocess Engineering:BBE
• /
• 제5권1호
• /
• pp.7-12
• /
• 2000

The expression of the mouse $\alpha-amylase$ gene in the methylotrophic yeast, P pastoris was investigated. The mouse $\alpha-amylase$ gene was inserted into the multi-cloning site of a Pichi a expression vector, pPIC9, yielding a new expression vector pME624. The plasmid pME624 was digested with SalI or BglII, and was introduced into P. pastoris strain GSl15 by the PEG1000 method. Fifty-three transformants were obtained by the transplacement of pME624 digested with SaiII or BglII into the HIS4locus $(38\;of\;Mut^+\;clone)$ or into the AOX1 locus $(15\;of\;Mut^s\;clone)$. Southern blot was carried out in 11 transformants, which showed that the mouse $\alpha-amylase$ gene was integrated into the Pichia chromosome. When the second screening was performed in shaker culture, transformant G2 showed the highest $\alpha-amylase$ activity, 290 units/ml after 3-day culture, among 53 transformants. When this expression level of the mouse $\alpha-amylase$ gene is compared with that in recombinant Saccharomyces cerevisiae harboring a plasmid encoding the same mouse $\alpha-amylase$ gene, the specific enzyme activity is eight fold higher than that of the recombinant S. cerevisiae.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권9호
• /
• pp.4587-4605
• /
• 2019

High Efficiency Video Coding (HEVC) suffers from high computational complexity due to its quad-tree structure in motion estimation (ME). This paper exposes an adaptive search range decision algorithm for accelerating HEVC integer-pel ME on GPU which estimates the optimal search range (SR) using a MAP (Maximum A Posteriori) estimator. There are three main contributions; First, we define the motion feature as the standard deviation of motion vector difference values in a CTU. Second, a MAP estimator is proposed, which theoretically estimates the motion feature of the current CTU using the motion feature of a temporally adjacent CTU and its SR without any data dependency. Thus, the SR for the current CTU is parallelly determined. Finally, the values of the prior distribution and the likelihood for each discretized motion feature are computed in advance and stored at a look-up table to further save the computational complexity. Experimental results show in conventional HEVC test sequences that the proposed algorithm can achieves high average time reductions without any subjective quality loss as well as with little BD-bitrate increase.

• 한국정보디스플레이학회:학술대회논문집
• /
• 한국정보디스플레이학회 2007년도 7th International Meeting on Information Display 제7권1호
• /
• pp.413-416
• /
• 2007

Motion-estimation/motion-compensation (ME/MC) provides superior motion picture quality but its huge computation load results in high cost. Impulsive driving is a cost-effective solution but it suffers from large flicker and brightness loss. Motion compensated impulsive driving technology has been developed to achieve high motion picture quality in a cost-effective implementation by combining ME/MC and impulsive driving. The key idea is to apply ME/MC or impulsive driving selectively according to the motion vector distribution of the incoming image sequence. In this paper, the description of the algorithm and the experimental results are provided.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제31권9호
• /
• pp.1142-1149
• /
• 2004

움직임 추정은 동영상내에 존재하는 중복된 데이타를 제거하기 때문에 비디오 영상 압축에서 중요한 역할을 하지만 높은 계산 복잡도로 인하여 실시간 영상 전송에 많은 어려움을 가지고 있다. 이러한 문제점을 해결하기 위해서는 낮은 계산 복잡도를 지닌 움직임 추정 기법들이 필요한데, 본 논문에서는 현재 블록과 높은 시간적, 공간적 상관성을 가지고 있는 블록들의 움직임 벡터들, 즉 참조 프레임에서 동일한 위치에 있는 블록의 움직임 벡터와 현재 프레임에서 현재 블록의 이웃에 있는 블록들의 움직임 벡터들을 이용하여 현재 블록의 탐색 시작점과 탐색 패턴을 적응적으로 결정하여 움직임 벡터를 추정하는 움직임 추정 기법을 제안한다. 실험을 통하여, 제안된 기법을 낮은 계산 복잡도를 지닌 움직임 벡터 필드 적응적 탐색 기법 (Motion Vector Field Adaptive Search Technique : MVFAST)과 예측된 움직임 벡터필드 적응적 탐색 기법(Predictive Motion Vector Field Adaptive Search Technique : PMVFAST)과 비교하였을 경우, 제안된 기법은 약 0.01～0.3 (dB) 정도의 화질 향상과 낮은 계산 복잡도로 인하여 약 1.12～l.33 배의 속도 향상을 보였다.

• 방송공학회논문지
• /
• 제16권2호
• /
• pp.293-304
• /
• 2011

본 논문에서는 H.264/AVC 인코더에서 가장 많은 연산 시간이 소요되는 움직임 추정(motion estimation, ME) 동작을 위한 하드웨어의 구조를 제안하고 IP(intellectual property) 형태로 구현하였다. 고속 움직임 추정기의 구조는 버퍼(buffer), PU 어레이(processing unit array), SAD 선택기(SAD selector), MV 생성기(motion vector generator) 등으로 구성되어 있다. PU 어레이는 16개의 PU로 구성되어 있고, 각각의 PU는 16개의 PE(processing element)로 이루어져 있다. 제안한 하드웨어의 동작적인 특징은 외부메모리 접근량을 줄이기 위해 현재와 참조프레임의 데이터를 재사용한다는 것과 SAD연산을 수행할 때 클록의 손실 없이 계산을 할 수 있다는 것이다. 구현한 고속 움직임 추정기는 Altera 사의 FPGA인 StatixIII EP3SE80F1152C2에서 3%의 자원을 사용하였고, 최대 동작주파수는 446.43MHz이었다. 따라서 구현한 하드웨어는 1080p 영상을 최대 50fps로 처리할 수 있다.

• Bulletin of the Korean Chemical Society
• /
• 제33권12호
• /
• pp.4161-4164
• /
• 2012

Density functional MO calculations for the methylsilole anion of $[C_4H_4SiMe]^-$ and methylgermole anion of $[C_4H_4SiMe]^-$ at the B3LYP (full)/6-311+$G^*$ level (GAUSSIAN 94) were carried out and characterized by frequency analysis. The ground state structure for the methylsilole anion and methylgermole anion is that the methyl group is pyramidalized with highly localized structure. The difference between the calculated $C_{\alpha}-C_{\beta}$ and $C_{\beta}-C_{\beta}$ distances are 9.4 and 11.5 pm, respectively. The E-Me vector forms an angle of $67.9^{\circ}$ and $78.2^{\circ}$ with the $C_4E$ plane, respectively. The optimized structures of the saddle point state for the methylsilole anion and methylgermole anion have been also found as a planar with highly delocalized structure. The optimized $C_{\alpha}-C_{\beta}$ and $C_{\beta}-C_{\beta}$ distances are nearly equal for both cases. The methyl group is located in the plane of $C_4E$ ring and the angle between the E-Me vector and the $C_4E$ plane for the methylsilole anion and methylgermole anion is $2.0^{\circ}$ and $2.3^{\circ}$, respectively. The energy difference between the ground state structure and the transition state structure is only 5.1 kcal $mol^{-1}$ for the methylsilole anion. However, the energy difference of the methylgermole anion is 14.9 kcal $mol^{-1}$, which is much higher than that for the corresponding methylsilole monoanion by 9.8 kcal $mol^{-1}$. Based on MO calculations, we suggest that the head-to-tail dimer compound, 4, result from [2+2] cycloaddition of silicon-carbon double bond character in the highly delocalized transition state of 1. However, the inversion barrier for the methylgermole anion is too high to dimerize.

• 방송공학회논문지
• /
• 제9권1호
• /
• pp.54-63
• /
• 2004

서로 다른 프레임 율을 가진 영상매체 사이의 프레임 호환을 위해, 움직임 추정 (motion estimation, ME)과 움직임 보상을 이용한 중간 영상 삽입 (합성) 기법 (motion compensated interpolation, MCI)이 활용되는데, 본 논문에서는 MCI에 적합한 움직임 추정기법을 제안하고, 제안한 움직임 추정 기법을 종래의 MCI에 적용한다. 종래의 방법에선 움직임 추정 블록과 MCI 블록의 크기가 동일하나, 본 논문에서 사용된 움직임 추정 블록은 MCI 블록과 같은 중심 축을 가지고, MCI 블록보다 더 크기 때문에 인접 블록과 중첩된 특성을 가진다. 제안한 움직임 추정 블록에 의한 계산량 증가를 줄이기 위해, 제안한 움직임 추정 블록내의 화소를 샘플링 하여 움직임 추정을 실행하였다. 제안한 방식을 검증하기 위해, 움직임 추정 블록을 다양한 샘플 계수로 샘플링 한 후 움직임 추정을 실행하였다. 제안한 방식으로 추정된 움직임 벡터 (motion vector, MV)를 활용하여 MCI를 수행하였으며, 수행된 결과를 종래의 방식에 의한 결과와 비교하였다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2005년도 추계종합학술대회
• /
• pp.639-642
• /
• 2005

Motion Estimation(ME) is an important part of video compression, because it requires a large amount of computation. Half-pixel and quarter-pixel motion estimation allows high video compression rates but it also has high computation complexity. In this paper we suggest a new and efficient motion estimation algorithm for half-pixel and quarter-pixel motion estimation using SAD values. In the method, an integer-pixel motion vector is found and then only three neighboring points of the integer-pixel motion vector is evaluated to find the half-pixel motion vector. The quarter-pixel motion vector is also found by using a similar method. Experimental results of our method shows 20% reduction in computation time, when compared with those of a conventional method, while producing same quality motion vectors.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 ITC-CSCC -2
• /
• pp.807-810
• /
• 2000

In this paper, a new hierarchical motion estimation (ME) scheme using the wavelet transformed multi-resolution image layers is proposed. While the coarse-to-fine (CtF) ME, used in previously proposed coding schemes, can provide a better estimate at the coarsest resolution, it is difficult to accurately track motion at finer resolution. On the other hand, in fine-to-coarse (FtC) ME, it can solves this local minima problem by estimating motion track at the finest subband and propagating the motion vector (MV) to coarser subband. But this method causes to higher computational overhead. This paper proposes a new method for reducing the computational overhead of fine-to-coarse rnulti-resolution motion estimation (MRME) at the finest resolution level by searching for the region to consider motion vectors of the coarsest resolution subband.