The Invention Relates to a Double Block Sleeper External Appearance Quality Automatic Inspection System
The double block concrete sleeper is the most used prefabricated components in Chinese high-speed railway construction.At present,it is mainly produced by assembly line prefabrication.The traditional manual inspection method cannot fully satisfy the requirements of the double block sleeper.In order to solve the problem of high accuracy requirements and heavy inspection tasks,this paper proposed a double block sleeper shape and appearance quality automatic inspection system based on high-precision 3D laser scanner and high-resolution industrial camera.The inspection system was integrated with production line,so that the inspection system was automated operated online with inspection function.This paper researched the automatic acquisition technology of 3D point cloud and high-definition image of double block sleeper,3D point cloud analysis and processing technology of double block sleeper,images analysis and processing technology of double block sleeper.Through point cloud segmentation and model reconstruction algorithm,the key components model of double block sleeper was obtained,automatic detection of the shape dimensions was realized.Through image template matching and image deep learning algorithm,intelligent recognition of sleeper numbers and cracks was realized.The inspect accuracy of the system can reach 0.1 mm,the detection efficiency is better than 3 minutes/group,the crack recognition accuracy is better than 95%,and the detection items and accuracy meet the relevant specifications.The system has the characteristics of automation,real-time,high precision,and high efficiency,can effectively solve the quality control problems of double block sleeper prefabricated factory
NETL
NSTL
万方数据
