【摘要】 针对静态的Bayer抖动算法会产生块状效应的缺陷,提出了随时间变换抖动矩阵的动态Bayer抖动算法.该改进算法主要根据人眼的视觉特性,对每帧视频图像套用轮换的Bayer抖动矩阵模板,通过多帧图像叠加,消除块状效应.用Verilog HDL对改进的抖动算法进行行为级描述,该图像抖动模块设计可支持1~4 bit抖动,通过仿真和综合,用FPGA进行了硬件实现.验证结果表明该动态抖动算法易于硬件实现,适用于动态图像的实时处理,并且在保持静态抖动算法优点的基础上,有效消除了块状效应.更多还原

【Abstract】 An improved algorithm with time-varying dithering matrix was proposed to avoid the blocky effect in the dithered images obtained from dynamic Bayer dithering algorithm.On the basis of human visual characters,the rotating Bayer dithering matrix template in every frame of images was used to average the blocky effect in visual effect through multiple frames.The improved arithmetic was described,simulated and synthetized with Verilog HDL.The design of this dithering module can support 1～4bit dithering processes.The field programmable gate array(FPGA) verification was also presented.The results showed that the dynamic dithering algorithm,being be implemented easy and adapted to the real-time process of dynamic images,can effectively remove the blocky effect,maintaining the advantages of static dithering algorithm.更多还原