Font Size: a A A

Technology Research Of On-line Defects Detection And Controlment In Screen Gap Based On Machine Vision

Posted on:2015-03-14Degree:MasterType:Thesis
Country:ChinaCandidate:W WuFull Text:PDF
GTID:2298330452958838Subject:Instrument Science and Technology
Abstract/Summary:PDF Full Text Request
In this paper, we developed a set of online automatic detection system based onmachine vision image processing technology in oil-pipe in laser cutting, throughinformation acquisition, image processing, camera calibration, distortion correctionand signal synchronization and other means, and ultimately realize the screensautomatic online slotted precision detection. The main research contents include:1We propose a detection scheme based on machine vision to build the model ofthe camera acquisition combined with on-site processing environment and thedetection accuracy requirements, we design image processing algorithms, and theimpact of the focus algorithms to analyze and improve accuracy, improve thedetection performance of the system.2In order to fully reduce errors to improve accuracy, the center of gravitymethod is proposed based on sub pixel kerfs edge, through the slotted edge coarsepositioning, the target region segmentation, and feature extraction algorithms such asgravity method points to achieve a slotted the sub-pixel edge detection to ensure theaccuracy of detection.3System analysis principles camera calibration and calibration accuracy andpropose a high-precision multi-camera calibration method of value Z. Through theprecise calibration target moving, it gets the relative position of the camera coordinateZ value. It targets the circle be calibrated to detect sub-pixel center. While detectingthe sub-pixel of calibration target in the center of the circle, and we use the Cannyalgorithm to detect coarse positioning edge circular target, then catch the coordinatesof the extracted pixels on the six boundary circle using least squares fit to obtain eachcalibration target board circular sub-pixel center coordinates, greatly improves theaccuracy of camera calibration.4Due to the information collection when the camera’s installation position andshooting angle, slotted image appears distorted phenomenon. Through the formationof distortion principle analysis, this paper presents a projection based on vanishingpoint perspective distortion correction methods. By calculating the coordinates of thevanishing point, respectively, of the X-direction and Y-direction distortion correctionis performed to achieve a good correction effect.5It builds a software module with the I/O to synchronize image acquisition cardcollection. By5V DC level signal produced when machining collection, it sets twotimers1and2in the program. The system uses Timer1to synchronize image acquisition, image processing using Timer2synchronized to achieve the image ofstability on line.
Keywords/Search Tags:Oil-piple, Machine Vision, Sub-pixel Edge Extraction, Vanishing Point, Synchronous Acquisition
PDF Full Text Request
Related items