Research On Image Correction Technology Based On Double Vanishing Point Perspective Transformation

Image correction refers to the process of restoring the tilt distortion or distorted image,which is widely used in safety monitoring,industrial manufacturing,intelligent transportation,logistics and other fields.With the development of imaging technology and correction methods,the application field is also expanding.Image correction has gradually become a research hotspot in the field of image processing.At present,most of the image correction methods are to extract feature points and simplify the camera model,so as to obtain better correction effect.However,these methods have high computational complexity and image restoration degree is difficult to meet the needs.Therefore,it is necessary to carry out the research of image correction methods to improve the correction efficiency and image restoration rate.The image rectification method based on double vanishing point is a good choice.The vanishing point information is extracted from the distorted image to reconstruct the projection matrix,to improve the correction efficiency and image restoration rate.This thesis has carried out the following research work:(1)Edge detection is an important part of image preprocessing.In this thesis,Canny operator is improved,and the edge detection method based on median filter and Gaussian filter and adaptive threshold segmentation is adopted.The experimental results show that the improved Canny operator can detect image edge efficiently.(2)In the distortion image vanishing point extraction,this thesis improves the random sampling consistency algorithm,proposes to select more than 4 pairs of matching feature points to establish a temporary model,according to the temporary model to remove the outer points in the sample,retain the inner points,and iteratively determine the best model to extract vanishing points.The experimental results show that the accuracy of the algorithm is better than the traditional random sampling consistency algorithm,and can accurately extract the vanishing point in the image.(3)This thesis corrects the image according to the information of image vanishing points,realizes the image correction technology based on double vanishing points,and processes the corrected image continuously to improve the quality of the corrected image.And through the evaluation index: running time,peak signal-to-noise ratio,image entropy,correction accuracy and the other two image correction methods are compared and analyzed in four scenes.The experimental results show that the correct rate of the image correction method based on double vanishing point is more than 95% in the four scenes of building image,license plate image,landscape image and daily sundry image,and the other three evaluation indexes are alsobetter than the other two correction methods.Therefore,the image correction method based on double vanishing point is a feasible method,which can supplement the research of image correction.