Track fastener bolt looseness automatic detection method based on 2D and 3D laser images

• Main Authors: LI LIN, CHEN WENTING, LUO WENTING
• Format: Patent
• Language: Chinese; English
• Published: 24.02.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention relates to a 2D and 3D laser image-based track fastener bolt looseness automatic detection method, which comprises the following steps of: acquiring a track fastener image by utilizing an independently researched and developed three-dimensional linear scanning laser imaging system, and training a bolt target identification model in a track fastener by adopting a YOLOv5 deep learning algorithm; classifying targets in the bolt detection frame into spikes and nuts according to the coordinate positions on the basis of a deep learning recognition result; three-dimensional depth information corresponding to the two types of targets is extracted, the height value mode of the three-dimensional depth information is counted, and the height difference is calculated; a threshold value is reasonably set, and finally automatic detection of bolt looseness is achieved. According to the invention, the bolt looseness can be automatically detected, the accuracy and efficiency of railway facility detection can be i