A void area identification method for georadar tunnel lining inspection images based on grayscale distribution

Tunnel lining voids are a common disease in tunnel engineering, posing a threat to the operational safety of tunnels. Traditional ground-penetrating radar (GPR) detection of tunnel lining images has low discrimination accuracy and cannot provide accurate detection results. This paper proposes a void area identification method for tunnel lining inspection images based on grayscale distribution. Firstly, four different models are constructed and detected using GPR to obtain the original detection images. Secondly, the original images are preprocessed using Reflex to achieve a recognizable effect. Finally, based on the feature analysis of the images, the images are enhanced through grayscale enhancement processing, edge detection, adaptive threshold segmentation, and the removal of small connected areas, resulting in clearer void areas with higher image accuracy.