海上拍摄图像的清晰化算法研究

Study on Clearness Algorithm of Image Captured at Sea

海上拍摄的图像易受到雾气与海面光影的影响,导致图像对比度低、可读性差,对后续图像分析造成了不利影响。论文提出了一种清晰化算法,改进了暗通道去雾方法中大气透射率的估计方法,利用引导滤波优化透射率,进而求解大气散射物理模型实现去雾。针对去雾后图像亮度较低,对比度差,不利于后续的图像分析问题,对去雾后图像进行gamma校正,增强对比度,改善可读性。将算法应用于典型图像,与经典算法相比,所提算法效果显著,实现了去雾功能,有效增强了目标对比度,图像自然且无失真情况,为进一步的图像分析打下了良好的基础。

The image captured at sea is vulnerable to the influence of fog and sea light and shadow, resulting in poor image contrast and poor readability, which has a negative impact on subsequent image analysis. A clearness algorithm is proposed in this paper, which improves the estimation method of atmospheric transmittance in dark channel dehaze method, and uses guided filter- ing to optimize transmittance, and then solves atmospheric scattering physical model to achieve fog removal. Aiming at the low brightness and contrast after dehazing, it is not conducive to subsequent image analysis. After the fog removal, gamma is corrected to enhance contrast and improve readability. The algorithm is applied to typical images. Compared with classical algorithm, the algo- rithm is effective, and achieves the function of fog removal. It effectively enhances the target contrast, and the image is natural and without distortion, which lays a good foundation for further image analysis.