提出一种基于改进颜色传递策略与非下采样 Contourlet变换(NSCT)的红外与可见光图像伪彩色融合算法。首先, 利用 NSCT与基于清晰区域的 Canny边缘检测算法获得灰度融合图。其次, 将融合灰度图像插入 Y通道, 源图像与融合灰度图像之间的残差图像分别插入 Cb、Cr通道以生成 YCbCr源彩色图像。最后, 利用本文设计的颜色传递模型对源彩色图像和目标图像进行色彩颜色统计匹配, 同时, 通过自适应颜色传递参数模型调整颜色传递参数。实验结果表明, 本文提出的融合算法使得伪彩色融合图像不仅对比度高、传递色彩自然、可以较好地抑制色彩渗入图像目标, 而且对目标图像质量要求不严格。

A pseudo-color fusion algorithm based on improved color transfer strategy and non down sampling Contourlet transform (NSCT) was presented. Gray fusion images were acquired by using NSCT and Canny edge detection algorithm based on clear region. YCbCr source color images were generated by inserting fusion gray level image and residual image respectively. In this paper, color transfer model is designed, it can be used to statistically match the source color image and the target image. Color transfer parameters were adjusted by an adaptive color transfer model. Experiments show that the proposed fusion algorithm not only makes pseudo-color fusion image high in contrast, but transfers the color naturally, contains the color penetration in the target image, and the quality requirement of the target image is not strict.