基于Laplace金字塔图像融合的含雾图像去雾方法

Haze Removal Method for Hazy Images Based on Laplacian Pyramid Image Fusion

摘要: 雾天场景中采集的图像易被大量雾气遮盖，导致图像清晰度大大降低、物体细节无法呈现，严重影响后续信息提取以及目标检测。针对这一问题，本文提出一种基于Laplace金字塔图像融合的去雾方法。首先将含雾彩色图像转换为灰度图像，分别采用传统小波去雾方法和直方图均衡化方法对灰度图像进行预处理，得到两幅中间处理图像，然后通过Laplace金字塔图像融合技术，生成去雾目标图像。实验结果显示，该方法在主观视觉上能更清晰还原景物细节信息，客观指标也明显优于文章提到的其他方法，表明该方法整体去雾效果更加理想，是一种行之有效的去雾方法。

Abstract: Images captured in foggy scenes are easily obscured by a large amount of fog, which significantly reduces image clarity, makes object details invisible, and seriously affects subsequent information extraction and target detection. To address this issue, this paper proposes a haze removal method based on Laplacian Pyramid image fusion. First, the hazy color image is converted into a grayscale image. The traditional wavelet haze removal method and histogram equalization method are respectively used to preprocess the grayscale image, resulting in two intermediate processed images. Then, the Laplacian Pyramid image fusion technology is applied to generate the haze-removed target image. Experimental results show that this method can more clearly restore the detailed information of scenes in terms of subjective vision, and its objective indicators are significantly better than other methods mentioned in the paper. This indicates that the proposed method has a more ideal overall haze removal effect and is an effective haze removal method.