Compressor Piston Roundness Error Digital Detection System Research

The rapid high-precision development of machinery industry,which puts forward new detection technology of parts. Particularly piston field with relatively high precision processing demand, the traditional manual detection method lags behind the way of modern machine production, therefore, developing low-cost, high-precision, high efficiency and non-contact piston machining error detection system has good practical significance for double-tube type linear electric piston compressor development, detection system based on image processing technology provides a quick and effective way, in this paper ,the author researches and develops piston detection system in selected lighting conditions, based on two-dimensional image information. The completion of major tasks, including the choice and analysis of test equipment, image noise reduction preprocessing, the image edge extraction, image distortion compensation processing and circle (ellipse) fitting methods such as the edge of research.Selects noise filter to process mages in order to eliminate the noise of obtaining the mage under natural condition,tests showed that the filter has good performance,The paper proposes that the composite Canny operation detects image edge, the test shows Canny operator can extract image edge effectively while removing noise. the paper makes distortion compensation correction on system considering the distortion of the camera case.The results show that the method is simple and easy to implement, and the compensation effect is good.Finally, studies circle and the ellipse fitting method, analyses and compares three kinds of advantages and disadvantages of fitting method, and put forwards improved rapid and effective highly circle center location algorithm based on their own characteristics. Programming experimental results show that the method can achieve better treatment effects; using the method of least squares to calculate errors of circle and ellipse fitting and analyzes error factors of the detection system.