基于提升小波的红外和可见光图像融合方法

Fusion for IR and visible images based on lifting wavelet transform

提出了一种基于提升小波变换的红外和可见光图像融合方法。对红外图像进行检测分割,将提取到的目标重要信息融合到可见光图像中。然后进行图像的提升小波分解,对不同尺度下小波系数进行融合,以像素的局部平均梯度为高频系数融合准则,充分加入原始图像的边缘细节信息。最后依据融合后的小波系数重构图像。实验结果表明,该方法改善了融合效果,提高了运算速度。

A new fusion method for IR and visible images based on lifting wavelet transform is presented. Firstly, the infrared image is segmented to extract the target regions, and the important target information is added to visible image. Then, the source images are decomposed using lifting wavelet transform respectively, and the coefficients are combined according to the rule based on local average gradient to add the details such as edges. Lastly, the image is reconstructed with fused coefficients. Experiment shows that the method is more effective in both computational speed and fused image quality than some traditional multi-resolution image fusion methods.