摘要：
It is significant in the aspect of improving the image compression and transmission to make data filtering for the target image.The target image data has high noise density.Thus,it makes the weighting matrix of PDE filter incomplete.Traditional filtering method mainly uses the amplitude information to work out the probability density function of data and builds the weighting matrix.It makes the weighting matrix data disordered.Therefore,it leads to big filtering error.In this paper,we propose an adaptive filtering method of data selectivity based on the variation PDE.Firstly,the continuous PDE is discretized,and the corresponding digital filter is generated by using the digital iterative method to make pretreatment to the original noisy data.Then,the amplitude information is used to obtain the density function of data noise and target measurement,and then the data detect threshold is given,and the likelihood ratio of amplitude information is calculated.Moreover,the weighting matrix of PDE filtering is built,and the filtering frequency response is obtained based on the variation PDE.Finally,the cost function of data selectivity adaptive filtering is acquired,and the selectivity adaptive filtering of target image data is completed.The simulation results show that the method has high signal to noise ratio.It can improve the robustness of data selectivity adaptive filtering effectively.%对目标图像进行数据滤波,在提高图像压缩及传输方面具有重要意义.由于目标图像数据噪声密度较大,使得PDE滤波器的加权矩阵不完整.传统的滤波方法主要利用幅值信息计算出数据的概率密度函数,建立加权矩阵,使得加权矩阵数据混乱,导致滤波误差大.提出基于变分PDE的数据选择性自适应滤波方法,首先将连续的PDE离散化,用数字迭代方法产生相应的数字滤波器,对原始噪声数据进行预处理,利用幅值信息获取数据噪声密度函数和目标量测密度函数,给出数据的检测门限,计算其幅值信息似然比,然后组建PDE滤波器的加权矩阵,得到基于变分PDE滤波器频率响应,获取数据选择性自适应滤波器的代价函数,在此基础上完成对目标图像数据选择性自适应滤波.仿真结果表明,所提方法信噪比高,可以有效地提升数据选择性自适应滤波的鲁棒性.
