Efficient Image Dehazing with Boundary Constraint and Contextual Regularization

项目主页：http://www.escience.cn/people/menggaofeng/Publication.html

摘要

提出的原因：Images captured in foggy weather conditions often suffer from bad visibility.

提出的方法：In this paper, we propose an efficient regularization method to remove hazes from a single input image. Our method benefits much from an exploration on the inherent boundary constraint on the transmission function. This constraint, combined with a weighted L1−norm based contextual regularization, is modeled into an optimization problem to estimate the unknown scene transmission. A quite efficient algorithm based on variable splitting is also presented to solve the problem.

方法优势：The proposed method requires only a few general assumptions and can restore a high-quality haze-free image with faithful colors and fine image details.

实验结果：Experimental results on a variety of haze images demonstrate the effectiveness and efficiency of the proposed method.

1. 现存的方法大多数研究方向为探索额外的先验和限制（priors和constrainst），跟随他们的脚步，提出

2. 方法：得益于对传输函数固有边界约束的探索，提出了一种有效的正则化方法来去除单输入图像中的模糊。该约束与基于L1-范数的加权上下文正则化相结合，被建模为一个优化问题来估计未知的场景传输。提出了一种基于变量分割的高效算法。

3. 贡献：

（1）a new constraint on the scene transmission：有很好的几何解释；

（2）a new contextual regularization：在去雾中加入滤波器组来减弱图像噪声，增强一些有趣的图像结构；

（3）an efficient optimization scheme,：快速处理大尺寸图像。

算法步骤：

 

 

 

实验参数：

正则化：

1，从模型修正上看，起了一个trade-off作用，用于平衡学习过程中两个基本量，名字诸如bias-variance、拟合能力-泛化能力、损失函数-推广能力、经验风险-结构风险等等；
2，从模型求解上看，正则化提供了一种唯一解的可能，众所周知，光用最小二乘拟合可能出现无数组解，加个L1或L2正则化项能有唯一解。