Based On Image Gray Gradient For The Research Of Roundness Error Evaluation

The precision of roundness error in workpiece cross section is one of the important indicator of evaluating the quality of the products,during the processing of workpiece machining with a circular cross section.With the continuous development of science and technology,the accuracy requirement of the roundness error of rising accordingly.It is so tedious and takes so much time that it’s difficult to meet the demand of real time in traditional roundness error evaluation and measurement.However,with advantages such as fast speed,high precision,the digital image processing has been used widely in the field of machining roundness testing.It’s a method of evaluating the roundness error based on digital image processing which using photoelectric converter(CCD),a machine can collect product images.But,it’s inevitable to take random noise which seriously affects the accuracy of detection,during the process of the detection system collect images in real time.So the accuracy of detection is affected seriously.Do non-maxima suppression of image’s gradient,and split image with the ideal threshold.Design the roundness evaluation algorithm.Prove the validity of the algorithm in this paper,by simulation round simulation experiment use Matlab.Gathering the images of He-Ne laser speckle,lens and annulus by machine vision system.And then research on roundness evaluation,verifie the feasibility of the algorithm in this paper.Specific work is as follows:(1)Introduces the basic knowledge of spatial filtering simply,and compared the influence of different gradient operator to detect image edge roundness.(2)According to the different influence to the image by gradient operator design a gradient algorithm.Set up two kinds of model of least square circle algorithm,extract the edge of simulation round and proceed roundness evaluation.(3)Set up two least squares model in the rectangular coordinate system and polar coordinates.To revise the solution of the nonlinear function in traditional least square.To solve the blindness of damp coefficient,initial value.Based on the theory of gradient,To improve the non-maxima suppression,segment the image by ideal threshold.Propose roundness evaluation algorithm based on gray gradient.Evaluation three magnitude roundness error of He-Ne laser speckle by this algorithm.After analyze the computational data,get this algorithm has higher edge location accuracy and smaller roundness error.(4)Manual measuring different diameters of lens and annulus by vernier caliper and turntable.Calculate the average of the data,as a reference standard.Collect the image of lens and annulus by the machine vision system.Then deal with the image by traditional algorithm and improved algorithm in this paper.After analyze the computational data,get improved algorithm in this paper has higher precision.