基于局部线性变换的色阶映射算法

Tone Mapping Algorithm Based on Local Linear Transformation

摘要:

为了提高高动态图像显示的视觉效果,提出了一种基于局部线性变换的色阶映射算法；针对图像直接线性压缩能力较差的缺点,本文提出一种局部的线性变换模型；由于局部窗的结构,是利用局部窗的中心像素点与窗内的其他像素点的均值,方差的关系构造而成,因此基于局部线性变换的色阶映射算法能够抑制由具有高对比度的图像边缘所引起的鬼影和光晕现象；首先,根据线性变换将图像的色阶映射转换为求解泛函的最优解；其次,通过推导求解出线性变换的参数并将泛函最优化问题转换为求解一个线性方程组问题；最后求解线性方程组,得出视觉效果较好的低动态范围图像；有效避免传统色阶映射出现的光晕和亮度不连续的现象,能够更多的保留高动态图像的细节；最后通过实验结果对比分析,通过信息熵和对比度两个参数可以体现文中算法较好。

Abstract:

In order to improve the visual effect of the high dynamic image display, a tone mapping algorithm based on local linear transformation is proposed. A local linear transformation model is proposed in this paper, which is based on the poor ability of direct linear compression of the image. Due to the structure of the local window, using the mean of other pixels within the local window of the central pixel point and window, the covariance structure, therefore based on local linear transformation of the color gamut mapping algorithm is able to suppress by having a high contrast image edge caused by ghosts and halo phenomenon. Firstly, according to the linear transformation will be the image color gamut mapping function to be solved for the optimal solution; secondly, through the derivation of a linear transformation parameters and functional optimization problem is transformed into solving a linear equation group; finally solving a linear system of equations that visual effects better low dynamic range images. Effectively avoid the phenomenon of halo and brightness of the traditional color gamut mapping is discontinuous, able to retain more high dynamic image details. Finally, through the comparative analysis of experimental results, the algorithm is better than other algorithms in this paper.