U-shaped welding seam detection method

The invention discloses a U-shaped welding seam detection method, which takes laser sensor data at a certain moment on a robot track route as prior information to carry out welding seam detection, andcomprises the following steps of: denoising data, namely denoising sensor data; determining a datum line to facilitate subsequent feature extraction; corner point detection: finding a welding seam inflection point and calculating the width of the welding seam; approximate area calculation: obtaiing the weld joint area through adoption of a differential method; weld joint height calculation: calculating the weld joint height by taking the datum line as a reference; performing inclination identification to obtain a central axis fitting straight line of the weld joint and calculate the angle between the central axis fitting straight line and the datum line; and contour detection: detecting the defects of the welding seam contour by using a deep learning method. According to the weld joint detection method, the defec