基于多散斑图组合-调制的鬼成像优化方法

利用多个散斑图组合的方式提出了一种新的鬼成像优化方法.利用桶探测器值的大小对照射到目标物体上的多个散斑图进行排序,使相邻散斑图间差异性减小;通过对相邻散斑图间的叠加以及对应桶探测器值的调制,有效降低待关联数据间冗余和数量,并利用不同的关联运算规则实现对目标物体的重构.数值仿真结果表明,对于目标图像,在总采样次数4000次、4个相邻散斑图组合方式下,所提方法与未组合-调制下传统鬼成像、差分鬼成像和正负调制鬼成像相比,其峰值信噪比/对比度分别提升了21.7%/27.3%、8.3%/17.8%和14.7%/25.7%;通过对20幅目标图像的数值模拟与结果分析,发现峰值信噪比/对比度提升率在15%和30%以上的占比分别为90%/85%和50%/55%,说明该方法具有较好的普适性和推广价值.

Technique and Technology in the Composition Mantra： Considerations Some Analytical Considerations

This paper relates to Stockhausen＇s composition method of the 1970s based on the concept called ＂formula＂. This method is in part a derivation and in part an improvement of serialism. Stockhausen applied it in his works starting from 1970, with the composition Mantra for two pianos and ring modulator, and refined it during his whole life. This paper will discuss the formula＇s concept and his application. In particular, the author tries to analyze the first composition where the formula＇s method is applied under two different points of view： the composition method and the rule of technology in the composition.