红外与可见光图像的自然感彩色融合能够显著提高人眼视觉的情景感知和目标探测能力。基于样本的融合算法是一种快速有效、实时性强的自然感彩色融合算法。针对已有算法构建颜色查找表时存在的样本数据冲突和查找表不完整两个难题，提出一种基于BP 神经网络的自然感彩色融合算法。算法采用BP 神经网络对图像样本的二维灰度向量(g1, g2)和三维色彩向量(R, G, B)进行非线性拟合，从而获得灰度与色彩间的映射关系f(g1, g2)→(R, G, B)，并由此构建完整的颜色查找表。融合时，由输入的双波段图像的灰度g1,g2 和颜色查找表得到彩色融合图像。实验结果表明，本文融合图像颜色自然，景物易于分辨，在清晰度、彩色性、映射准确性方面已经达到甚至优于Toet 算法的图像融合效果。

The natural color fusion of infrared images and visible images can significantly improve the situation perception and target detection of human vision. Sample-based color fusion is a fast, effective, and real-time natural color fusion algorithm. In view of the problems of the existing algorithm in terms of the color look-up table construction, i.e., sample data collision and look-up table incompleteness, this paper proposes a new natural color fusion algorithm based on the BP neural network. In the algorithm, the mapping f(g1, g2)→(R, G, B) between grayscale and color is obtained by using the BP neural network to nonlinearly fit between the two-dimensional grayscale vector (g1, g2) and the three-dimensional color vector (R, G, B) of the image simples. Subsequently, the color look-up table is constructed based on the mapping. During color fusing, the fused image is obtained by the color look-up table and the input grayscale g1,g2 of dual-band images. The experiments show that the fused images based on the proposed algorithm have natural colors and are easily distinguishable objects. The fusion results obtained by the proposed algorithm are almost as good as or even better than the fusion results by Toet’s method in terms of definition, colorfulness, and mapping accuracy.