基于非线性直方图变换的对比度畸变图像校正

在实际成像过程中,因受多种因素影响而导致图像对比度畸变,使其质量下降。为了改善图像质量并增强视觉效果,提出了一种基于非线性直方图变换与参数优化的降质图像对比度畸变校正方法。首先,针对常规直方图均衡方法的局限性,通过人眼视觉感知特性分析,引入图像直方图的先验约束条件,建立了非线性直方图变换模型;进而,从校正效果的最优化角度考虑,运用遗传算法的进化寻优进行校正参数的优化,从而形成了一种高性能的对比度畸变校正算法。一系列在不同场景实拍的对比度严重畸变图像的校正实验结果表明,本文方法校正结果的客观质量评价测算指标和实际视觉效果都有显著提升,和常规校正方法相比具有明显优势。

Correction of contrast distortion image based on nonlinear transform of histogram

In the practical imaging, the image contrast is distorted due to influence of multi-factors, which degrades image quality greatly. A contrast correction method is presented based on nonlinear transform of histogram and parameter optimization in order to improve image quality and enhance the visual effect. Firstly, in view of the limitation of conventional histogram equalization method, a nonlinear transform model of histogram was established based on the prior constraints of image histogram through the analysis of human visual perception characteristics. Then the parameters of transform model are optimized with the evolutionary searching of genetic algorithms to achieve optimal corrective effect so that a high performance contrast distortion correction algorithm emerged. A series of correction experimental results for real images with severe contrast distortion from different scenes demonstrate that the proposed algorithm outperforms the conventional correction methods in both of objective quality assessment and subjective visual effect and is provided with distinct advantages.