基于大气散射模型的红外图像增强方法

An Infrared Image Enhancement Method Based on the Atmospheric Scattering Model

为更有效地提升红外图像的整体视觉效果并恢复出其中的场景目标,提出了一个基于大气散射模型的红外图像增强方法。首先,对红外图像进行反转操作,并引入大气散射模型对其降质机理进行描述。然后,利用四叉树分解技术将图像分割为一系列子块,并提出相应的基于子块的模型因子估计策略来恢复出场景目标。最后,结合导向全变分模型和一种基于Retinex模型的修正算法,进一步提高增强后图像的视觉效果。主观及客观对比实验结果证明了本算法具有良好的鲁棒性,及在视觉效果增强、有效信息增益方面的优势。

An infrared image enhancement method based on the atmospheric scattering model is proposed,which aims at improving the global visibility as well as increasing the scene details.The infrared image is converted into a reversed one,and the corresponding degradation mechanism is described based on the atmospheric scattering model.Then,the target image is segmented into a set of blocks using the quad-tree decomposition,and therefore the scene objects can be recovered within each blocks using the proposed model coefficients estimation strategies.Then,the guided total variation model as well as a Retinex-based algorithm are introduced to further improve the visual effect of the enhanced image.The comparison experimental results verify that the proposed method can produce results comparative to and even better than several existing state-of-the-art techniques with respect of the robustness and effectiveness.