【摘要】 提出一种新的图像融合方法,该方法首先利用小波变换对要融合的图像进行多尺度小波分解;然后对分解后的各层上不同频带的子图像采用不同的融合处理技术,即在低频域内采用平均算子进行融合处理,以保留图像的背景信息,在高频域内对分解后的小波系数矩阵采用区域特征进行融合处理;最后将融合的低频分量和高频细节分量结合进行小波逆变换得到融合图像。采用该融合算法对两幅不同位置聚焦的图像进行融合实验,结果表明,采用该方法可以得到较好的融合效果。更多还原

【Abstract】 A new method of image fusion was put forward.First,original image was decomposed into many subimages by the multiscale decomposition of wavelet transform.Second,different fusing technologies were adopted for the subimages with different frequency bands.The fusion method of av- erage operator was adopted for the subimages with low-frequency band in order to reserve the back- ground information of image.The fusion method of regional features based on wavelet coefficient ma- trices was adopted for the subimages with high-frequency bands.Last,the whole fused image could be obtained by inverse wavelet transform of the low-frequency and the high-frequency subimages which were processed by the above proposed methods.An experiment of two images with different fo- cus locations with the proposed method of image fusion was given,and it indicates that the better fu- sion image can be obtained by the proposed method.更多还原