论文部分内容阅读
推广Lawton的方法构造FIR的复数子波包变换滤波器组及相应的子波包函数,并按照不同的时间 频率特性选择适用的基函数.针对传统子波和子波包分解系数难以准确反映该时频单元信号分量强度的缺点,给出改进的自适应子波包变换技术——群归一子波包变换,并利用时频掩模滤波技术,完成强背景下目标信息的获取与分离,明显改善信杂比.由于该类复子波包变换滤波器是线性相位的酉正交滤波器,分解系数去相关能力更好,处理效果更强.推广的l~p-范数熵定义增加了群归一子波包变换在低信杂比情况下的处理能力.类似的结果可应用于强散射点分离、图象增强、杂波抑制、最优检测以及雷达成象等信号处理领域.
The method of Lawton is extended to construct the FIR complex wavelet packet transform filter bank and the corresponding wavelet packet function, and the suitable basis function is selected according to different time-frequency characteristics. In view of the fact that the traditional wavelet and wavelet packet decomposition coefficients can not accurately reflect the time Frequency signal strength, the improved adaptive wavelet packet transform technology - group-normalized wavelet packet transform, and the use of time-frequency mask filtering technology to complete the strong background of target information acquisition and separation, significantly Improve signal-to-noise ratio.As the complex wavelet packet transform filter is a unitary orthogonal filter with linear phase, the decomposition coefficient has better de-correlation ability and stronger processing effect. The extended l ~ p-norm entropy definition increases Group-normalized wavelet packet transform in low signal-to-noise ratio (SNR), similar results can be applied in the field of signal processing such as strong scattering point separation, image enhancement, clutter suppression, optimal detection and radar imaging.