Research And Development Of Roundness And Cylindrical Measurement For Revolving Based On Computer Vision

In industry the shaft is a very common machine parts. Roundness and cylindrical are one of the important form error for shaft. They are also an important parameter to evaluate the parts and components. The existence of integrated form error directly affecting the parts and components'accuracy, rotation accuracy, vibration, noise etc, and reduceing their life.With the development of modern precision, ultra-precision machining technology, modern error theory, electronic technology, the rotational parts has made a inceasingly higher requirements of circular and cylindrical error. Currently, in industrial application, the large quantities of products'roundness and cylindrical is detected by humen, the measurement accuracy is low. The high accuracy instrumentations are very expensive and demanding on the operating environment, which can't meets the online, real-time, and flexible requirements of modern measurement techniques.Machine Vision has the advantages of efficient, flexible,accurate and easy to implement automation. The main topic of this issue is to achieve the shaft's roundness and cylindrical testing using machine vision, in order to meet the needs of online measurement.Firstly the article summarizes and outlines the roundness and cylindrical's pacificat-ion theory, and the measurement,methods used in measuring.Secondly, by analyzing the main causes of error in image detection, proposing a method which established the formal equation of error equation using standard parts. Image distortion is corrected and calibrated at the same time the coefficient of normal-type is calculated. This method is not only simple, but also but also effective to improve the accuracy of the image detection.Thirdly, based on the traditional V-measurements, analyzed the reflecting process of roundness in the V-shaped grooves. Proposing a method which established a error equation by image edge detection and realized the detection of roundness error.Finally, based on the roundness, by decomposition in to roundness, straightness and parallelism, achieved the cylindrical testing in computer vision.The engine valve's roundness and cylindrical data were got through experiment, which compared with the results of roundness, cylindrical instruments. Proved the feasibility of the proposed method.