Key Technologies Of Vision Positioning System Of Wire Bonder

In the welding process of the wire bonder,the positioning accuracy of the visual positioning system has a direct influence on the welding quality.In order to achieve high-quality and high-efficiency wire bonding,a high-precision,high-speed and high-stability visual positioning system is required.In order to improve the visual positioning accuracy of the wire bonder,and improve the production efficiency and product quality,a research on the key technology of the visual positioning system of the wire bonder is proposed.The study include the establishment of the visual positioning system,the visual positioning algorithm,the quality assessment method of the collected images,and the simple detection method of the optical lens of the wire bonder.According to requirements of the vision positioning system of the wire bonder,the hardware of the camera,lens and light source is carried out,and the structure of the vision system is designed.According to the requirements of the vision positioning system of positioning accuracy,efficiency and stability,a template matching algorithm is proposed,which uses normalized product correlation to calculate the similarity value of the image.The pyramid technology is used to speed up the matching time of the image.The method of quadric surface fitting is used to obtain the positioning accuracy of the image,and the method of combining multi-angle template and fan-shaped vector is used to match the image angle.For the problem of quality of collected image,the image quality assessment based on the generative adversarial network is introduced into the visual positioning system.In the research of image quality assessment technology,the pseudo-reference image of the collected pictures is generated by the generative adversarial network,which converts the non-parameter image quality assessment into pseudo-parameter image quality assessment.Aiming at the test requirements of the image acquisition system,a simple optical lens detection method is proposed,which can effectively,conveniently and quickly detect the magnification,uniformity and resolution.Through experiments,the effectiveness of the visual positioning algorithm and image quality assessment algorithm are proposed.At the same time,the solder joints were inspected for quality and tensile and thrust testing to prove that the entire visual positioning system can meet the final use requirements.