Research On Measurement System Of Red Hot Workpiece Size Based On SD-OCT

Non-contact measurement techniques have advanced significantly in recent years in the area of dimensional inspection for hot workpieces.When it comes to quick measuring speed and broad applicability,optical or laser-based technologies have distinguished themselves as having a distinct advantage.However,getting crucial details such edges and corners for self-emitting red-hot workpieces is difficult,which restricts the advancement of accuracy in optical or laser-based measurement methods.As a result,many businesses still use manual methods to measure the dimensions of hot,high-temperature workpieces or they wait for the workpiece to cool before measuring it.These difficulties worsen the issue since it takes too long to measure the dimensions of hot,high-temperature workpieces,which leads to product waste and a loss of raw materials and production.This research suggests a red-hot workpiece dimensional inspection method based on Spectral Domain Optical Coherence Tomography(SD-OCT)to overcome the difficulties in dimensional inspection of red-hot workpieces.The dimensional inspection of red-hot workpieces is made possible because the excited light from the hot object cannot interfere with the laser of the OCT system.The dimensional inspection system ensures a specific degree of accuracy while expanding the measuring range by utilizing the high resolution advantage of OCT and incorporating a motion platform to overcome the constrained imaging depth range.Real-time dimensional inspection of hot workpieces is made possible by the suggested technology,giving crucial data support for process development and quality control in hot workpiece applications.First,this study used SD-OCT technology to achieve dimensional inspection of hot workpieces.An OCT module and a motion platform were created using theoretical analysis.A dimensional inspection system was built in order to enable the measurement of red-hot workpiece dimensions by carefully considering the system’s axial resolution,lateral resolution,and image depth.Furthermore,to collect surface position information for the workpiece,a position information extraction approach based on tomographic imaging was also suggested.First,a multi-structure fusion morphological filter was developed to improve OCT tomographic imaging quality by removing image noise while retaining contour edge details.Then,to simplify threshold modification,an adaptive threshold-based Canny edge detection approach was presented for edge detection and extraction of straight lines in tomographic imaging.Last but not least,by specifying edge feature points and accumulator filtering rules,optimizing Hough detection computing resources,and achieving the extraction of surface position information of the workpiece,improvements based on Hough line detection were made.Additionally,the module division was carried out in accordance with the dimensional inspection system’s functional requirements.A Queue Message Handler(QMH)framework was built using the layered architecture idea for effective message queue processing.The system enabled parallel execution of tasks like data acquisition,processing,and storage,feedback control of the motion platform using a Proportional-Integral-Derivative(PID)controller,and the execution of the position information extraction algorithm based on tomographic imaging by implementing multiple threads.The software for the dimensional inspection system was developed with the goal of enhancing the system’s real-time performance,scalability,and reusability.Finally,experimental validation was carried out using the created high-temperature red-hot workpiece dimensions inspection platform and testing software.The test specimens’dimensions were measured in the cooling,high temperature,and room temperature stages.The studies showed that dimensional inspection of hot workpieces could be accomplished with the suggested SD-OCT-based dimensional inspection system,the enhanced QMH-based testing software,and the tomographic imaging-based position information extraction algorithm.