Conversion of underwater concrete images to air in detection of hydraulic structures

Abstract

Taking underwater concrete images with an optical camera is an important measure for underwater defect detection. However, the underwater low-light environment and light refraction on the surface of different media result in poor image quality. In order to make the collected underwater images effectively reflect the real situation of underwater concrete defects, we propose an image conversion algorithm combining underwater image color enhancement and refraction distortion correction, which can convert underwater images into aerial equivalent images. In this paper, two conversion models of underwater image to air conversion are proposed, one of which models the refractive distortion of the underwater multilayered media through the ray projection method to correct the refractive distortion of underwater images and image field of view (FOV) conversion. The other is that by analyzing the problem of low image-imaging quality and loss of edge information due to the uneven illumination environment underwater, we process the dark channel priority adaptive enhancement algorithm to improve underwater image quality. The experimental results from real scenes show that the imaging quality of underwater images is improved by converting underwater images to air. The pixel error of images converted into the air is ⩽1 pixel, and the FOV error of images is ⩽8.5%. The high-precision underwater image conversion provides strong support for subsequent underwater measurement.