基于偏振成像的水下退化图像复原算法

Underwater Degraded Image Restoration Algorithm Based on Polarization Imaging

采用当前方法对水下退化图像进行复原时,不能有效的消除噪声对水下退化图像复原过程产生的影响,复原后水下退化图像的信息熵较低,存在复原效率低和图像质量低的问题。为此提出基于偏振成像的水下退化图像复原算法。构建图像退化模型,采用高斯核函数在退化模型的基础上对水下退化图像进行去噪处理。根据去噪结果建立水下偏振成像模型,估计水体透射率和全局背景光,根据估计值实现水下退化图像的复原。实验结果表明,与当前方法相比,所提算法的复原效率更高,信息熵更高,说明上述算法处理后的图像质量更好,有效解决了当前方法存在的问题,使用价值较高。

In current methods,the influence of noise on the restoration process of underwater degraded image cannot be effectively eliminated.The information entropy of underwater degraded image after restoration is low,lead-ing to low restoration efficiency and low image quality.Therefore,an algorithm of underwater degradation image res-toration based on polarization imaging was proposed.A model of image degradation was constructed.Based on this model,the Gaussian kernel function was used to remove the noise from the underwater degradation image.According to the denoising results,an underwater polarized imaging model was built.Moreover,the water transmissivity and global background light were estimated.According to the estimated value,the underwater degraded image was re-stored.Simulation results show that the proposed algorithm has higher recovery efficiency and information entropy.The image quality processed by the algorithm is better,which effectively solves the problems of current method,so it has good application value.