Abstract:Integer motion estimation (IME), which acts as a key component in video encoder, is to remove temporal redundancies by searching the best integer motion vectors for dynamic partition blocks in a macro-block (MB). Huge memory bandwidth requirements and unbearable computational resource demanding are two key bottlenecks in IME engine design, especially for large search window (SW) cases. In this paper, a three-level pipelined VLSI architecture design is proposed, where efficiently integrates the reference data sharing search (RDSS) into multi-resolution motion estimation algorithm (MMEA). First, a hardware-friendly MMEA algorithm is mapped into three-level pipelined architecture with neglected coding quality loss. Second, sub-sampled RDSS coupled with Level C + are adopted to reduce on-chip memory and bandwidth at the coarsest and middle level. Data sharing between IME and fractional motion estimation (FME) is achieved by loading only a local predictive SW at the finest level. Finally, the three levels are parallelized and pipelined to guarantee the gradual refinement of MMEA and the hardware utilization. Experimental results show that the proposed architecture can reach a good balance among complexity, on-chip memory, bandwidth, and the data flow regularity. Only 320 processing elements (PE) within 550 cycles are required for IME search, where the SW is set to 256 × 256. Our architecture can achieve 1080P@30 fps real-time processing at the working frequency of 134.6 MHz, with 135 K gates and 8.93 KB on-chip memory.

A Scalable Frame-Level Pipelined Architecture for FSBM Motion Estimation

A Hardware-Efficient VLSI Architecture for Variable Block Size Motion Estimation in H.264/AVC

Architecture Design for H.264/AVC Integer Motion Estimation with Minimum Memory Bandwidth

A New Architecture for H.264 Variable Block Size Motion Estimation

Fast Multiframe Selection Algorithm Based on Spatial-Temporal Characteristics of Motion Field.

Fast Multi-frame Motion Estimation Algorithms for H.264

VLSI Architecture for H.264 Integer-Pel Motion Estimation with Minimum Memory Bandwidth Requirement

An Adaptive And Fast H.264 Multi-Frame Selection Algorithm Based On Information From Previous Searches

A High Parallel-Pipelined Real-Time Estimation Architecture for MPEG-4 Simple Profile

Software Pipelines Design for Variable Block-Size Motion Estimation with Large Search Range

An Adaptive and Fast Fractional Pixel Search Algorithm in H.264

Improved Ffsbm Algorithm And Its Vlsi Architecture For Variable Block Size Motion Estimation Of H.264

VLSI Oriented Fast Multiple Reference Frame Motion Estimation Algorithm for H.264/AVC

AN ADAPTIVEAND FASTH.264MULTI-FRAMESELECTIONALGORITHM BASED ON INFORMATION FROM PREVIOUSSEARCHES

Scalable VLSI Architecture for Variable Block Size Integer Motion Estimation in H.264/AVC.

A New Fast Motion Estimation Algorithm Based on H.264

Area-efficient Implementation of Low Bit-Depth Motion Estimation Algorithms

Design of Motion Estimation Algorithm Based on Pipeline FPGA

An Efficient Fractional Motion Estimation Architecture for Avs Real-Time Full Hd Video Encoder

Three-level Pipelined Multi-Resolution Integer Motion Estimation Engine with Optimized Reference Data Sharing Search for AVS.

Fast motion estimation algorithm based on edge block detection and motion vector information