A research on detecting and recognizing bridge cracks in complex underwater conditions

The method aims to recognize and extract the characteristic parameters of bridge cracks based on images of the cracks obtained through the application of preprocessing technologies, such as graying, graphical enhancements, spatial filtering, gray-level threshold segmentation, etc.. The approach has been tested for accuracy to avoid the incorrect identification of chaff as a method error. The proposed method has proved to be rather accurate and effective in extracting information on cracks from the bridge image tests.