Dynamic Reconfigurable Structure with Rate Distortion Optimization

The Rate Distortion Optimization(RDO)algorithm in High Efficiency Video Coding(HEVC)has many iterations and a large number of calculations.In order to decrease the calculation time and meet the requirements of fast switching of RDO algorithms of different scales,an RDO dynamic reconfigurable structure is proposed.First,the Quantization Parameter(QP)and bit rate values were loaded through an H⁃tree Configurable Network(HCN),and the execution status of the array was detected in real time.When the switching request of the RDO algorithm was detected,the corresponding configuration information was delivered.This self⁃reconfiguration implementation method improved the flexibility and utilization of hardware.Experimental results show that when the control bit width was only increased by 31.25%,the designed configuration network could increase the number of controllable processing units by 32 times,and the execution cycle was 50%lower than the same type of design.Compared with previous RDO algorithm,the RDO algorithm implemented on the reconfigurable array based on the configuration network had an average operating frequency increase of 12.5%and an area reduction of 56.4%.

Parallelization of motion compensation algorithm based on reconfigurable video array processor

The new encoding tools of high efficiency video coding(HEVC) make the interpolation operation more complex in motion compensation(MC) for better video compression, but impose higher requirements on the computational efficiency and control logic of the hardware architecture. The reconfigurable array processor can take into consideration both the computational efficiency and flexible switching of algorithms very well. Through mining the data dependency and parallelism among interpolation operation, this paper presents a parallelization method based on the dynamic reconfigurable array processor proposed by the project team. The number of pixels loaded from the external memory is reduced significantly, by multiplexing the common data in the previous reference block and the current reference block. Flexible switching of variable block operation is realized by using dynamic reconfiguration mechanism. A 16×16 processor element(PE)’s array is used to dynamically process a 4×4-64×64 block size. The experimental results show that, the reference block update speed is increased by 39.9%. In the case of an array size of 16 PEs, the number of pixels processed in parallel reaches 16.