Research Of Image Correction Technology In Vision Measuring System

Vision measuring systems are widely used to detecte shape and measure the dimension of object in industrial field.They use the images to reconstruct the3D geometric information. If the image distortion is exited, it must cause big measurement error. Because of shotting angle of the camera, tilt distortion will appear, and which is difficult for computer to extract and recognize the object features. Meanwhile, because of the lens manufacture precision’s influence, geometry distortion will appear in images such as varying of geometry position, size, shape, position, and this will reduce the measurement accuracy of vision measurement systems. So before of shape detection and dimension measurement, we must correct the tilt distortion and the geometry distortion.The processing of image distortion correction in Vision measuring system is as follow:first we correct the tilt distortion, then we correct the geometry distortion to get the undistorted images, then we turn the non-integral coordinate into integral pixel coordinate by using the gray interpolation technique, fillally we finish the image correction. Main tasks in this paper are as follows:(1) The categorys of tilt distortion are described, the method based on Hoff transform and Projection method are analyzed.A efficient K-means Clustering Mothod is presented in this paper.(2) The causation and category of geometry distortion are described, the classical distortion correction methods such as Tsai Method,Weng’s Method, Zhang’s Method and Geometric Invariance Calibration Methods are analyzed.A improved anti-interference mothod,named geometric distortion correction method based on uniformly-spaced lines,is presented in this paper.(3) The Efficiency and effect of three different gray interpolation techniques such as the Nearest Neighbor Interpolation, Bilinear Interpolation and Cubic Convolution Interpolation is studied, some useful advices for choosing a appropriate interpolation method for different vision measuring system are given in this paper.Finally, the image correction mothod presented in this paper is transplanted in our Express Automatic Categorization System.The result shows that my method can obviously correct the geometry distortion and tilt distortion, and improve the measurement accuracy of Express Automatic Categorization System.