The invention discloses a
monorail track finger-shaped plate defect detection method and
system, and relates to the technical field of finger-shaped plate defect detection.The
monorail track finger-shaped plate defect detection method comprises the steps that image data of a track surface is collected, preprocessed and then input into a
deep learning target detection model, and a finger-shaped plate rectangular frame area and a finger-shaped plate category are extracted; segmenting the rectangular frame region of the finger-shaped plate, and extracting a finger-shaped plate region; opening operation is carried out on the finger-shaped plate area, the number of fingers in the finger area is extracted, and if the number of the fingers is not met, it is judged that
finger detection of the finger-shaped plate is abnormal; based on the finger region of the finger-shaped plate, measuring the
shortest distance between a single finger and the root of an opposite finger, and if the
shortest distance exceeds the length of the corresponding finger, determining that the finger of the finger-shaped plate is disengaged; and performing closed operation on a finger region of the finger-shaped plate, and selecting a corresponding mode to judge finger collision by judging whether an overlapping region exists or not. According to the invention, the image data is combined with the
deep learning technology to judge whether the adjacent finger-shaped plates have detection abnormity, finger separation and finger collision, so that intelligent track inspection is realized, and the inspection efficiency is improved.