保细节的水下欠曝光图像的复原和增强

RESTORATION AND ENHANCEMENT OF UNDERWATER UNDER-EXPOSURE IMAGES WITH DETAIL PRESERVING

因水下环境复杂、光线差、颜色退化等问题,提出一种保护边缘细节并增强图像颜色的水下欠曝光图像的复原和增强方法。采用改进的小波阈值函数去噪,色彩恢复的多尺度视网膜增强(MSRCR)联合改进引导滤波器的方法去雾;运用自适应曝光图方法改善欠曝光图像的亮度;利用深度学习融合改进的自适应伽马校正技术解决颜色退化和边缘弱化的问题。实验表明,相比于其他的方法,该方法具有更好的视觉效果。

Due to the complex underwater environment,poor light,and color degradation,this paper proposes a method for restoration and enhancement of underwater underexposed images that protects edge details and enhances image color.An improved wavelet threshold function was used to denoise,and the multi-scale retinex color restoration(MSRCR)combined guided filtering method was used to remove haze;adaptive exposure maps were used to enhance the brightness of underexposed images;deep learning algorithms were combined with the improved auto Gamma correction technology to solve the problems of color degradation and edge weakening.Experiments show that this method has better visual effects than other methods.