论文部分内容阅读
提出了基于一种新的小波具有紧支撑、正交性、伸缩矩阵为 2 00 2 的非张量积小波的图像融合方法。首先根据非张量积小波理论 ,利用Daubechies构造的单变量滤波器构造出基于四通道的不可分的小波滤波器组 ,用此滤波器组对参加融合的图像进行分解 ,然后对低频部分采用取均值、高频部分采用系数绝对值取大的融合算法对分解子图进行融合 ,最后重构。并采用熵、交叉熵、互信息、均方根误差和峰值信噪比等指标对该方法进行了客观评价。对可见光图像与红外图像、远红外图像与近红外图像、遥感图像、多聚焦图像和其它多类图像的融合实验结果证明本方法有较好的融合效果 ,其融合性能与采用同样融合算法的张量积db2小波的融合方法的融合性能相当。
An image fusion method based on a new wavelet with tight support, orthogonality and non-tensor product wavelet with scaling of 2 00 2 is proposed. Firstly, based on the theory of non-tensor product wavelet, a univariate filter based on Daubechies is constructed to construct an irreducible wavelet filter bank based on four-channel. The filter bank is used to decompose the images which participate in the fusion. Then, , High-frequency part of the coefficient of the absolute value of the fusion algorithm to decompose the sub-image fusion, the final reconstruction. The method is objectively evaluated by entropy, cross entropy, mutual information, root mean square error and peak signal to noise ratio. Experimental results on the fusion of visible and infrared images, far-infrared and near-infrared images, remote sensing images, multi-focus images and other multi-class images demonstrate that the proposed method has a better fusion effect, and its fusion performance is similar to that of the same fusion algorithm The fusion performance of the product db2 wavelet fusion method is equivalent.