一种保边缘的图像去噪方法

提出一种保边缘的去噪方法。该方法利用4个方向核卷积检测图像中的噪声点和非噪声点,对非噪声点不进行处理。在噪声点的3×3邻域中选择距离最小的非噪声点,若无非噪声点则邻域扩大为5×5,若有非噪声点则用这些点的中值替换噪声点,否则用邻域中距离最远像素点的均值替换噪声点。实验结果表明,该方法不仅能有效去除图像的噪声,且能较好地保护图像边缘。

一种基于小波变换的混合数据校正法

采用一种将小波变换方法与移动中值滤波方法相结合的混合方法来校正数据.通过仿真和结果对比,表明该方法在过程测量数据同时含有高斯噪声和非高斯噪声(离群点)的情况下具有较好的鲁棒性.

Asymmetric Double Quantum Dots as a Subminiature Mesoscopic Cell

A subminiature mesoscopic cell, consisting of asymmetric double quantum dots capacitively coupled to a nearby mesoscopic circuit, is proposed, which can transform disordered noise energy to ordered electric energy. Two schemes, the noises originating from the nearby mesoscopic circuit and from the electromagnetic wave disturbance in external environment, are investigated. We found that the proposed cell can manifest as a good constant current source and the output current may not reach its largest value even if the circuit is shorted.