Rate-Distortion and Complexity Optimized Motion Estimation for H.264 Video Coding

Abstract

H.264 video coding standard supports several interprediction coding modes that use macroblock partitions with variable block sizes. Rate-distortion optimal selection of both the motion vectors and the coding mode of each macroblock is essential for an H.264 encoder to achieve superior coding efficiency. Unfortunately, searching for optimal motion vectors of each possible subblock incurs a heavy computational cost. In this paper, in order to reduce the computational burden of integer-pel motion estimation without sacrificing from the coding performance, we propose a rate-distortion and complexity joint optimization framework. Within this framework, we develop a simple method that determines for each macroblock which partitions are likely to be optimal. Motion vector search is carried out for only the selected partitions, thus reducing the complexity of the motion estimation step. The mode selection criteria is based on a measure of spatio-temporal activity within the macroblock. The procedure minimizes the coding loss at a given level of computational complexity either for the full video sequence or for each single frame. For the latter case, the algorithm provides a tight upper bound on the worst case complexity/execution time of the motion estimation module. Simulation results show that the algorithm speeds up integer-pel motion estimation by a factor of up to 40 with less than 0.2 dB loss in coding efficiency.

Mode Decision-Based Algorithm for Complexity Control in H.264/AVC

“…In [14], an estimation of the high frequency content of a block and a target complexity are included in a novel cost function so that the ME process relies on it to decide which partitions are taken into account for each MB. In [15], modified versions of both J motion and J mode cost functions are proposed by adding a complexity term that is based on the computation time and the number of instructions required.…”

“…Although many research efforts have been devoted to reduce the complexity of video compression algorithms [1]- [13], only a few works have been devoted to actually control the complexity [14]- [24]. In this paper, the problem of complexity control is tackled in the framework of H.264/AVC, the latest video coding standard of the Joint Video Team (JVT).…”

Complexity scalable motion estimation control for H.264/AVC

“…COMPLEXITY SCALABLE ME ME is computationally very intensive, taking up to 60-80% of total encoding time [6]. Therefore, our focus is on making complexity scalable ME.…”

Real-time H.264 video encoding in software with fast mode decision and dynamic complexity control

“…A method that limits the reference frame and mode search based on the predicted complexity and an assessment of the benefits of search depth is described in [Wu et al 2007]. [Ates and Altunbasak 2008] describes a method that limits mode search based on a spatiotemporal activity metric and uses a Lagrangian parameter for trading off complexity and RD performance. Unlike the work herein, these last two methods do not incorporate forward SKIP prediction or a frame buffering technique.…”

“…The lack of a frame buffer means that the frame encoding time budget must be set conservatively. A graph in [Ates and Altunbasak 2008] shows that the lack of a frame buffer results in 20% of clock cycles going unused. [Kaminsky et al 2008] proposes a method that provides Constant Bit Rate (CBR) and constant complexity encoding.…”