Research On Image Edge Distorrtion Mechanism And Correction Technology In Visual Measurement System

Image distortion processing is an important research content of image processing technology and visual measurement technology.According to the different distortion conditions of image,selecting the corresponding distortion processing method to improve the image has great application value to improve the accuracy of visual measurement technology,and it is one of the key research contents in the direction of medical image and industrial image distortion at home and abroad at present.Aiming at the image edge distortion caused by light scattering in visual measurement,this thesis proposes an image gray anomaly correction algorithm and an image edge spot elimination method based on light polarization,and adjusts the mechanical structure of the imaging system to compensate for the position error.A visual measurement system consisting of an imaging system,a lighting system and a light polarization device was used for experimental verification.The main work of this thesis includes the following contents:(1)By referring to the existing research data on image edge distortion at home and abroad,combined with the digital image processing technology and the main characteristics of image distortion,the mechanism of image edge distortion and the main influencing factors are analyzed,and the image correction technology for different types of distortion is studied.According to the image acquisition requirements of the visual measurement system,the experimental platform was built,and the selection scheme was made for the imaging device,lighting system,image acquisition card and the overall mechanical structure of the measurement system.At the same time,the appropriate camera imaging scheme was selected for the acquisition of subsequent images.(2)The image gray anomaly correction algorithm is studied and verified by experiment.Firstly,the acquired image is denoised by Gaussian filter function,sobel operator is used to extract the pixel-level edge of the image,and gray moment algorithm is used to locate the subpixel edge of the image.After the accurate image edge is obtained,the abnormal gray value point generated by distortion on the image edge is located by the anomaly correction algorithm,and the gray change curve of the anomaly point is obtained.The peak region generated by abnormal mutation in the gray curve of the anomaly point is located by the peak location function,and the segmentation Hermite interpolation is used to interpolate and replace the gray data in the region.To achieve the effect of anomaly correction.(3)The surface of the measured object with high reflectance is more conducive to the experiment.Therefore,aiming at this condition,the shims with smooth surface are selected as the experimental object,and the appropriate polarizer and fixed bracket are selected based on the imaging system.The polarization extinction experiment is completed based on the polarization effect of light.Shim images with different polarization angles and light intensity were collected by experiments,and the gray change of the same edge point in each image was analyzed to verify the experimental correction effect.Finally,the position error introduced by the imaging system is compensated by mechanical debugging to improve its impact on the image.