彩色融合图像在场景深度感知、目标探测识别、减少判断时间等方面优于灰度融合图像, 重点研究了红外与可见光彩色图像融合算法。针对经典的自然感图像融合算法中计算复杂度高、缺乏颜色恒常性问题, 提出基于 YUV颜色空间恒参数颜色传递( CPCT)图像融合算法, 并通过栅栏、山丘、森林和城镇四组实验验证了本文算法在颜色恒常性和算法实时性上有较大改善。

Color fusion images are better than gray fusion images in scene perception, reducing estimate time, target detection and recognition. Hence, the thesis mainly studied the infrared and visible color image fusion algorithms. To solve the problem of complicated computation and no color constancy in the classical color transfer based image fusion method, a color image fusion method is proposed that is based on constant parameters color transfer in YUV color space. The experimental results show that this method can greatly improve the real-time performance with color constancy.