基于小波-Contourlet变换的偏振图像融合算法

针对传统图像的信息保留不充分,以及偏振图像妨碍视觉观察、纹理细节不理想等问题,提出了一种基于小波-Contourlet变换(WBCT)的偏振图像融合算法.首先,将预处理过的4幅偏振角度图像,通过Stokes方法得到偏振强度图像和偏振度图像;后采用WBCT变换分解,低频系数采用PCA变换方法进行融合,高频系数通过区域特性能量的融合规则进行融合,最终的偏振融合图像由WBCT逆变换高低系数获得.实验结果表明,在主观视觉上,图像观察舒适性较好;并且,通过选取方法的对比,融合后的图像在客观评价指标上,皆优于选取的方法.

Polarizing Image Fusion Algorithm Based on Wavelet-Contourlet Transform

Traditional polarization image fusion is notorious for poor visual effects and insufficient retention of texture details.To address these problems,a new method of polarization image fusion using wavelet-based contourlet transform(WBCT)is proposed.Initially,the polarization intensity image and polarization degree image are obtained by the Stokes method for the 4 pre-processed polarization angle images.After decomposition by WBCT transform,the low-frequency coefficient is fused using the PCA transform method,and the high-frequency coefficient is fused using the regional characteristic energy fusion rule.Finally,the fused high-and low-frequency coefficients are converted by WBCT inverse transform to obtain a final fusion image with polarization characteristics.Results indicate that the image observation is comfortable in subjective vision.Moreover,comparison of the selected methods reveals an improvement in the evaluation data of the fused images in terms of objective evaluation indices.