基于多核异构的异源图像融合抗晕光系统设计

Design of heterogeneous image fusion anti-halo system based on multi-core heterogeneity

针对异源图像融合抗晕光算法复杂,应用于视频实时处理困难,设计了基于嵌入式多核异构架构的异源图像融合抗晕光系统。将系统控制和图像融合抗晕光算法分别在ARM和FPGA中实现,解决了数据串行和任务顺序执行导致的系统实时性问题,消除了融合图像中的晕光痕迹。经实验验证,多核异构的抗晕光系统晕光消除度和系统实时性均满足系统需求。

Aiming at the complex anti-halo algorithm of heterologous image fusion and the difficulty of real-time video processing,a heterogeneous image fusion anti-halo system based on embedded multi-core heterogeneous architecture is designed.The system control and image fusion anti-halo algorithm were implemented in ARM and FPGA respectively,which solved the real-time problem caused by data serial and task sequence execution,and eliminated the halo trace in the fusion image.The experimental results show that the halo elimination and real-time performance of the anti-halo system can meet the requirements of the system.