Research On Contour Reconstruction Of Conical Thread In Image Detection

In the industrial production life, thread is common part, it also play an important and an irreplaceable role in fastening parts, connecting parts, sealing parts or any other places. A product can use or not based on the quality of the thread is good or not. People are very concerned about the problem of how to ensure the quality of the thread. It is very meaningful that exacted and rapid detection the parameters of thread when we the thread is done. Early detection methods are manual measurement, people use thread gauge or universal tools to detect the thread, called contact measurement. The way of contact measurement is high costly, heavy workload, low efficiency, accuracy of the measurement result is difficult to protect, and the measurement tools and test thread connected is likely to cause the surfaced of the tools and thread wear and tear. It affected the test result.As image processing technology based on vision has been improved, applying image processing in work piece detection is widely recognized; Techniques and devices of image detection thread is also being constantly improved and perfected. It is a non-contact detection method; we get the image from the CCD camera, and obtain the parameter information what we need after the image processing. It improves the thread detection accuracy, reduce the loss of tools and reduce the error of manual. However, the thread image analysis and image correction of the conventional detection technique are not perfect.During the detection of the taper thread based on image, the pint is to obtain the image of the tooth profile of the taper thread accurately. But the image cannot reflect the actual information of the tooth after the primal image processing of taper thread by existing method, the only result we can obtain is the profile of thread image. The paper analyzed the reason of errors between the taper thread image and the tooth profile, revealed the geometric relationship between different locations of tooth generatrixwhen helix rising and thread image profiles, deduced the equation of various points on the tooth generatrix.And establishing the boundary equation for the theoretical image profile, then get differential equation of the theoretical tooth profile. By using the image processing module in Labview combined with the mathematical analysis in Matlab which reconstituted the thread image profile, we can extract the parameters of actual tooth profile and achieve a more accurate thread detection. The paper also set up a system for thread detection and obtained the real thread image, using the new method to reconstruct the thread profile and extract the parameters of actual thread profile, then compared them with the results detected by traditional method. So the new method can improve the precision of measurement over the taper thread based on image.