AMLCD运动伪像仿真及优化

Motion blur simulation and optimization for AMLCD

为了高效准确预估液晶显示器的动态成像质量,提出了一种通用的多参数AMLCD彩色自然图像运动伪像仿真算法.该算法以显示系统时序分析和液晶响应经验公式为基础,利用视频内容数据,对每帧内人眼追踪路径上各个子像素的亮度进行积分,在不同仿真条件下均能准确地模拟主观感受.利用该算法对高速液晶材料、倍频和背光调制3种目前改善AMLCD运动模糊的主流技术进行了对比.结果表明,三者中背光调制技术改善效果最显著,但也容易产生伪轮廓重影效应,必须严格保证背光的同步性并慎重选择调制的占空比.某手机制造商根据上述结论预估了不同条件下多种液晶屏的性能,最终产品性能与预估完全吻合,验证了算法的正确性与实用性.

To pre-estimate the dynamic image quality of AMLCD （active matrix liquid crystal dis-play）efficiently and precisely,a multi-parameter simulation algorithm is proposed for the motion blur artifact of colorful natural images.Based on the time sequential analysis of the display system and the empirical formula of liquid crystal response curves,the light impulse of every sub pixel is in-tegrated within each frame along the eye trace using data from the video content.Thus,the subjec-tive perception of human eyes is simulated accurately under a variety of conditions.Using this algo-rithm,three current mainstream techniques for suppressing motion blur including faster response time,double refresh rate and blinking backlight are compared.Results show that among these three methods,blinking backlight provides the best result.However,it also brings ghost artifact unless the backlight synchronization is perfect and the duty cycle is selected carefully.According to the conclu-sions above,some cell phone manufacturer pre-estimated the performance of several AMLCDs under different conditions.It is found that the estimation fits well with the actual performance of the prod-ucts,which verifies the correctness and practicability of the algorithm.