一种基于总变分与显著性检测的红外与可见光图像融合方法

红外与可见光图像融合是多源信息融合中的一个重要研究内容,它在军事侦察等方面有着广泛的应用.本文基于总变分模型和显著性检测方法,提出了一种有效的融合方法.首先,通过对红外与可见光图像的特征分布考察,构建了一个信息融合的总变分模型.其次,基于亮度对比度的显著性检测,给出了总变分模型中保真项权值的估计方法.实验仿真表明,无论是视觉观察还是客观评价,本文的方法均比一些现有方法体现了更好的结果.

Total variation and saliency detection based infrared and visible image fusion

As an important aspect in multi-source information fusion, infrared and visible image fusion has been widely applied in many fields such as military reconnaissance. Based on a total variation model and a saliency detection-based method, this paper proposes an effective fusion method. First, through investigating the feature distribution of infrared and visible images, a total variational model is constructed for our fusion task. Moreover, for the fidelity weights of the total variation, an estimation method is given, based on the saliency detection with luminance contrast(LC). Experimental results show that the proposed method is superior to other methods in both visual observation and objective evaluation.