[PDF] Top 20 Algorithm and Architecture Design of Bandwidth-Oriented Motion Estimation for Real-Time Mobile Video Applications

... constant and sufficient memory bandwidth, regardless of the limited and dynamic bandwidth resources available for power-limited mobile ...that, for D1 (704 576) ... See full document

... pattern and a variable data BW ...amount of reference block data for the same SR ± ...constant and fixed during ME operations, which in turn assumes that the available BW is sufficient ... See full document

... path of the hardware for global elimination algorithm (GEA) is too long to meet the real-time constraints for high-end ...GEA and its corresponding ...groups and ... See full document

... Terms—H.264/AVC, motion estimation, video ...flexible and powerful with the development of semiconduc- tors, digital signal processing, and communication technology, in which the ... See full document

... data for motion estimation. In hardware design, [18], [19] proposed the hardware architec- tures to reduce the memory bandwidth overhead by adopting the binary motion ... See full document

... Low-Power and Bandwidth-Efficient Motion Estimation IP Core Design Using Binary Search Shih-Hao Wang, Shih-Hsin Tai, and Tihao Chiang, Senior Member, IEEE Abstract— A new ... See full document

... PROCEDURE OF FME IN H.264 Figure.1 shows the inter prediction flow of a 16x16 macro-block (MB) for fractional motion estimation (FME) in ...decision algorithm, the Lagrangian mode ... See full document

... Architecture Design of Belief Propagation for Real-Time Disparity Estimation Yu-Cheng Tseng, Student Member, IEEE, and Tian-Sheuan Chang, Senior Member, IEEE ... See full document

... GEA and its architecture design for fast block matching ...Our algorithm can solve several problems en- countered by ...block, and then to precisely compare the best roughly ... See full document

... (二) 以 8 顆 SPE 平行運算進行加速 我們使用 8 顆 SPE 平行運算的方式與前面 提到那篇的做法不同，他的方法是將每張參考 影像放在不同的 SPE 裡面運算，而我們的做法 則是將原本的影像拆成八分，如圖 5.(a)；與一 顆 SPE 計算一張影像比較，拿 8 顆 SPE 把一張 影像平分的作法在 DMA 傳輸的資料量就有差 別了，在一顆算一張影像的情況下，每顆 SPE 必須取得 2 張完整的影像，而拆成八份的作法 ... See full document

... dx and dy. Add motion vector dx and dy to the estimated complex motion parameters. After proposed the combinative algorithm, the performance comparison of the propose[r] ... See full document

... designed for the previous stan- dards, where VBS and MRF are not supported, the parameter of our design is set as the single-iteration 4SS with one reference ...processes and supply ... See full document

... Descent algorithm. In this algorithm, 91.2% memory bandwidth and 80% iterations can be ...impact of the irregular memory access is largely reduced by the proposed Reference-Based ... See full document

... The simulation results show that for the threshold voltage variation (AVTH), the constant current the proposed circuit can compensate for AVTH and improve the.. source is difficult to de[r] ... See full document

... Hierarchical Motion Estimation Algorithm and Its VLSI Architecture Bing-Fei Wu, Senior Member, IEEE, Hsin-Yuan Peng, Student Member, IEEE, and Tung-Lung Yu Abstract—This paper ... See full document

... We the integer-pel MVs and both the reference and current frames take MBrow x MBCo1 pixels as a macroblock (MB) and let as inputs to find the smallest SADs and the corresponding MVs the [r] ... See full document

... Fractional Motion Estima- tion (FME) with rate-distortion constrained mode decision can improve the rate-distortion efficiency by 2–6 dB in peak signal-to-noise ...must for real-time ... See full document

... Furthermore, if the PERFORMANCE OF CANDIDATE LEVEL DR FOR THE PROPOSED PARALLEL power is very low, we can even support only half-pel or integer- ARCHITECTURE.. pel resolution with one re[r] ... See full document

... suitable for software and hardware implementation. For hardware design comparison, the proposed algorithm is simple to be implemented in hardware and similar to the full search ... See full document

... Tree architecture, all distortions of a searching candidate are generated in the same cycle, and by an adder tree, distortions are accumulated to derive the SAD in one ...utilization and data ... See full document