Detect and identify earth rock embankment leakage based on UAV visible and infrared images

Abstract

Leakage is one of the most common and harmful disease in earth rock river embankment. Timely detection and accurate identification of the leakage in flood season is of great significance to ensure the safety of river embankment. Existing detection technologies generally have the defects of large blind area, low efficiency and high operation cost. And as a result, manual visual observation is still the main means to find embankment leakage in flood season. In order to reduce the huge labor resources consumption and provide an efficient and cheap method for embankment leakage detection in flood season, a non-contact sensing mode of UAV equipped with visible and infrared cameras is introduced into river embankment leakage detection. A simple and efficient leakage identification method based on UAV infrared images is also proposed. Considering the limitations of single vision, a leakage region segmentation method based on infrared and visible image fusion is proposed. The application of the proposed methods is demonstrated on an actual river embankment, and the results show that the proposed strategy is feasible to identify embankment leakage and it has a great application prospect for embankment leakage detection in flood season.