排水管道图像雾气及光斑干扰去除方法研究

近年来，管道内窥摄像检测是常用的排水管道缺陷检测技术，由于大部分管道内部具有一定污水充满度，管道图像往往存在白雾以及光斑等干扰，影响图像缺陷的识别和解译，为此，提出了一种排水管道图像雾气及光斑干扰去除方法，首先结合管道雾气特征，构建暗通道先验算法去雾，针对去雾后的色偏问题，使用一种多尺度Retinex色偏恢复算法，最后改进二维伽马函数，通过降低镜头污染，实现对图像光斑的压制。实验结果表明，该方法能有效去除排水管道内视频图像中的雾气和光斑的干扰问题，提高排水管道图像的可视度、对比度和清晰度，从而提高排水管道病害解译准确度。

Research on removing fog and spot interference in drainage pipeline image

In recent years, pipeline endoscopic camera detection is a common drainage pipeline defect detection technology. Because most pipelines have a certain sewage filling degree, there are often white fog and spot interference in pipeline images, which affect the identification and interpretation of image defects. Therefore, a method for removing fog and spot interference in drainage pipeline images is proposed, Firstly, combined with the characteristics of pipeline fog, the dark channel a priori method(DCP) is used to remove the fog. Aiming at the problem of color deviation after defogging, a multi-scale Retinex color deviation recovery algorithm is used. Finally, the two-dimensional gamma function is improved to suppress the image spot by reducing lens pollution. The experimental results show that this method can effectively remove the interference of fog and light spot in the video image of drainage pipeline, improve the visibility, contrast and clarity of drainage pipeline image, and improve the accuracy of drainage pipeline disease interpretation.