基于方向可调的不可分离小波框架变换图像融合检测方法

Image Fusion Detection Based on Adjustable and Non-separable Wavelet Frame Transform

针对复杂的中长波红外图像的特点,论文提出了一种基于方向可调的不可分离小波框架变换红外图像融合检测方法。首先,对图像进行不可分离小波变换分解,提取两幅图像的低频和高频,分别对高频和低频信息进行融合,并用逆变换重构成融合后的图像。对融合后的图像进行目标检测得出最终结果。实验结果表明文中提出的方法优于仅用长波或中波红外图像的目标检测。

According to the features of mid/long-wave infrared image ,this paper puts forward a fusion detection meth-od based on wavelet frames transform .First ,Low-frequency and high-frequency information are extracted from original mid/long-wave image using wavelet frames .High-frequency and low-frequency information fusion independently ,and reconstruct fused images by wavelet frames transform .The final target detection of the fused images obtain the final result .This paper shows that this method is better than using only long wave or medium wave infrared image detecting targets .