Research On The Key Technologies Of Vision Detection In Workpiece Flaw

Machine vision inspection system is widely applied to workpiece detection having advantages of high speed, accuracy, and non-contact,which is an effective way to overcome the shortage of artificial inspection and increase the degree of automation in the workpiece production. The workpiece flaw detection is an important content in the area of intelligentized workpiece detection. The main content and innovation of this paper are as follows:Firstly, this paper analyses the performance of current image acquisition system, constructs image acquisition experimental platform and achieves real-time, high-quality collection of the workpiece image.Secondly, according to the geometric regularity of the image, this paper proposes a method of image processing based on NSCT (Nonsubsampled Contourlet Transform). On the basis of multi-resolution decomposition, with semi-soft threshold this paper achieves denoising the image, making signal to noise ratio of the image significantly increased. Combining NSCT and modulus maxima to detect image edge, it extracts smooth edge and shows its ability of strong noise immunity.Thirdly, this paper studies detection and identification of workpiece defection. On the basis of Hough transform circle detection method targeting the circle center, we use template matching to realize top bottle defect detection and then identify workpiece's defect types through measuring parameters of defect's perimeter, area, elongation and circularity.Fourthly, we verify feasibility of DSP image processing algorithm through lots of experiments, develop the image acquisition and processing program and transplant the system software from PC to DSP. Continuous real-time detection of dynamic images is discussed and friendly user interface is designed.