• Title/Summary/Keyword: Fast Motion estimation algorithm

Search Result 286, Processing Time 0.03 seconds

A Fast Motion Estimation Algorithm with Motion Analysis (움직임 해석을 통한 고속 움직임 예측 알고리즘)

  • Jun, Young-Hyun;Yun, Jong-Ho;Cho, Hwa-Hyun;Choi, Myung-Ryul
    • Proceedings of the IEEK Conference
    • /
    • 2005.11a
    • /
    • pp.339-342
    • /
    • 2005
  • We present an efficient block-based motion estimation algorithm with motion analysis. The motion analysis determines a size of search pattern and a maximum repeated count of search pattern. In case of large movement in large image, we reduce search points and the local minimum which caused by low performance. The proposed algorithm employs with searching step of 2. The first step determines an initial search point with neighbor block vector and a size of initial search pattern. The second step determines a size of search pattern and a maximum repeated count with motion analysis. We improve motion prediction accuracy while reducing required computational complexity compared to other fast block-based motion estimation algorithms.

  • PDF

Fast Motion Estimation Algorithm via Minimum Error for Each Step (단계별 최소에러를 통한 고속 움직임 예측 알고리즘)

  • Kim, Jong Nam
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.20 no.8
    • /
    • pp.1531-1536
    • /
    • 2016
  • In this paper, we propose a fast motion estimation algorithm which is important in performance of video encoding. Even though so many fast algorithms for motion estimation have been published due to its tremendous computational amount of for full search algorithm, efforts for reducing computations in motion estimation still remain. In the paper, we propose an algorithm that reduces unnecessary computations only, while keeping prediction quality the same as that of the full search. The proposed algorithm does not calculate block matching error for each candidate at once to find motion vectors but divides the calculation procedure into several steps and calculates partial sum of block errors. By doing that, we can estimate the minimum error point early and get the enhancement of calculation speed by reducing unnecessary computations. The proposed algorithm uses smaller computations than conventional fast search algorithms with the same prediction quality as full search.

Efficient Fast Motion Estimation algorithm and Image Segmentation For Low-bit-rate Video Coding (저 전송율 비디오 부호화를 위한 효율적인 고속 움직임추정 알고리즘과 영상 분할기법)

  • 이병석;한수영;이동규;이두수
    • Proceedings of the IEEK Conference
    • /
    • 2001.06d
    • /
    • pp.211-214
    • /
    • 2001
  • This paper presents an efficient fast motion estimation algorithm and image segmentation method for low bit-rate coding. First, with region split information, the algorithm splits the image having homogeneous and semantic regions like face and semantic regions in image. Then, in these regions, We find the motion vector using adaptive search window adjustment. Additionally, with this new segment based fast motion estimation, we reduce blocking artifacts by intensively coding our interesting region(face or arm) in input image. The simulation results show the improvement in coding performance and image quality.

  • PDF

270 MHz Full HD H.264/AVC High Profile Encoder with Shared Multibank Memory-Based Fast Motion Estimation

  • Lee, Suk-Ho;Park, Seong-Mo;Park, Jong-Won
    • ETRI Journal
    • /
    • v.31 no.6
    • /
    • 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.

Fast Variable-size Block Matching Algorithm for Motion Estimation Based on Bit-pattern (비트패턴을 기반으로 한 고속의 적응적 가변 블록 움직임 예측 알고리즘)

  • 신동식;안재형
    • Journal of Korea Multimedia Society
    • /
    • v.3 no.4
    • /
    • pp.372-379
    • /
    • 2000
  • In this paper, we propose a fast variable-size block matching algorithm for motion estimation based on bit-pattern. Motion estimation in the proposed algorithm is performed after the representation of image sequence is transformed 8bit pixel values into 1bit ones depending on the mean value of search block, which brings a short searching time by reducing the computational complexity. Moreover, adaptive searching methods according to the motion information of the block make the procedure of motion estimation efficient by eliminating an unnecessary searching of low motion block and deepening a searching procedure in high motion block. Experimental results show that the proposed algorithm provides better performance-0.5dB PSNR improvement-than full search block matching algorithm with a fixed block size.

  • PDF

Fast Multi-Reference Frame Motion Estimation Algorithm Using a Relation of Motion Vector with Distance of Each Reference Frame (움직임 벡터와 참조 프레임간의 거리를 이용한 고속 다중 참조 프레임 움직임 추정)

  • Byun, Ju-Won;Choi, Jin-Ha;Kim, Jae-Seok
    • Journal of the Institute of Electronics Engineers of Korea SP
    • /
    • v.47 no.2
    • /
    • pp.69-76
    • /
    • 2010
  • This paper proposed a new fast multi-reference frame motion estimation algorithm. The proposed algorithm reduces search areas of motion estimation using a linear relation of motion vector with distance of each reference frame. New algorithm executes full search area motion estimation in reference frame 0 and reference frame 1. After that, search areas in reference frame 2, reference frame 3 and reference frame 4 are minimized by distance of each reference frame and results of motion estimation in reference frame 0 and reference frame 1. The proposed algorithm does not use a threshold value which is obstacle of hardware implementation and processing time schedule. Also, it reduced computation quantity of multi-reference motion estimation. Hardware implementation of multi-reference frame motion estimation is possible by these features. Simulation results show that PSNR drop and bitrate increase of proposed algorithm are lower than those of previous fast multi-reference frame motion estimation algorithm The number of computation of new algorithm is reduced 52.5% and quality of result is negligible when compared with full search area motion estimation which has 5 reference frames.

Modified Cross Search Algorithm for Fast Block Matching Motion Estimation (고속 블록 정합 움직임 추정을 위한 개선된 교차 탐색 알고리즘)

  • Ko, Byung-Kwan;Kwak, Tong-Ill;Hwang, Bo-Hyun;Yun, Jong-Ho;Choi, Myung-Ryul
    • Proceedings of the IEEK Conference
    • /
    • 2008.06a
    • /
    • pp.811-812
    • /
    • 2008
  • In this paper, a modified cross search algorithm for fast block matching motion estimation is proposed. Various Motion Estimation (ME) algorithms have been proposed since ME requires large computational complexity. The proposed algorithm employs Modified Cross Search Pattern (MCSP) to search the motion vector. Efficient compression can be achieved since Modified Cross Search Algorithm (MCSA) simplifies the search pattern to reduce the computational complexity. The experimental results show that proposed algorithm reduces the search points up to 29% compared to conventional methods.

  • PDF

Fast Motion Estimation Algorithm Based on Thresholds with Controllable Computation (계산량 제어가 가능한 문턱치 기반 고속 움직임 예측 알고리즘)

  • Kim, Jong-Nam
    • Journal of the Institute of Convergence Signal Processing
    • /
    • v.20 no.2
    • /
    • pp.84-90
    • /
    • 2019
  • Tremendous computation of full search or lossless motion estimation algorithms for video coding has led development of many fast motion estimation algorithms. We still need proper control of computation and prediction quality. In the paper, we suggest an algorithm that reduces computation effectively and controls computational amount and prediction quality, while keeping prediction quality as almost the same as that of the full search. The proposed algorithm uses multiple thresholds for partial block sum and times of counting unchanged minimum position for each step. It also calculates the partial block matching error, removes impossible candidates early, implements fast motion estimation by comparing times of keeping the position of minimum error for each step, and controls prediction quality and computation easily by adjusting the thresholds. The proposed algorithm can be combined with conventional fast motion estimation algorithms as well as by itself, further reduce computation while keeping the prediction quality as almost same as the algorithms, and prove it in the experimental results.

Design of a Fast Multi-Reference Frame Integer Motion Estimator for H.264/AVC

  • Byun, Juwon;Kim, Jaeseok
    • JSTS:Journal of Semiconductor Technology and Science
    • /
    • v.13 no.5
    • /
    • pp.430-442
    • /
    • 2013
  • This paper presents a fast multi-reference frame integer motion estimator for H.264/AVC. The proposed system uses the previously proposed fast multi-reference frame algorithm. The previously proposed algorithm executes a full search area motion estimation in reference frames 0 and 1. After that, the search areas of motion estimation in reference frames 2, 3 and 4 are minimized by a linear relationship between the motion vector and the distances from the current frame to the reference frames. For hardware implementation, the modified algorithm optimizes the search area, reduces the overlapping search area and modifies a division equation. Because the search area is reduced, the amount of computation is reduced by 58.7%. In experimental results, the modified algorithm shows an increase of bit-rate in 0.36% when compared with the five reference frame standard. The pipeline structure and the memory controller are also adopted for real-time video encoding. The proposed system is implemented using 0.13 um CMOS technology, and the gate count is 1089K with 6.50 KB of internal SRAM. It can encode a Full HD video ($1920{\times}1080P@30Hz$) in real-time at a 135 MHz clock speed with 5 reference frames.

A Past Elimination Algorithm of Impossible Candidate Vectors Using Matching Scan Method in Motion Estimation of Full Search (전영역 탐색 방식의 움직임 예측에서 매칭 스캔 방법을 이용한 불가능한 후보 벡터의 고속 제거 알고리즘)

  • Kim Jone-Nam
    • Journal of Korea Multimedia Society
    • /
    • v.8 no.8
    • /
    • pp.1080-1087
    • /
    • 2005
  • Significant computations for full search (FS) motion estimation have been a big obstacle in real-time video coding and recent MPEG-4 AVC (advanced video coding) standard requires much more computations than conventional MPEG-2 for motion estimation. To reduce an amount of computation of full search (FS) algorithm for fast motion estimation, we propose a new and fast matching algorithm without any degradation of predicted images like the conventional FS. The computational reduction without any degradation in predicted image comes from fast elimination of impossible candidate motion vectors. We obtain faster elimination of inappropriate motion vectors using efficient matching units from localization of complex area in image data and dithering order based matching scan. Our algorithm reduces about $30\%$ of computations for block matching error compared with the conventional partial distortion elimination (PDE) algorithm, and our algorithm will be useful in real-time video coding applications using MPEG-4 AVC or MPEG-2.

  • PDF