Research On High-precision Three-dimensional Detection And Recognition Method Of Pin Coplanarity

In order to ensure the quality of the mounting in the SMT production process,the coplanarity of the pins of the chip should have a high precision.The traditional contact detection method not only has the problem of low detection efficiency,but also often damage the chip pin because it has to contact the chip pins when it is detected.Therefore,it can not meet the high efficiency,high precision and low cost detection requirements of current chip detection.In this dissertation,a set of non-contact three-dimensional detection system for the coplanar of chip pins for SOP packaging is developed.The main contents of this dissertation are as follows:(1)After analyzing and testing demand,we choose the appropriate line laser sensor and two-dimensional displacement platform to Abuild a high-precision 3D scanning system to get the 3D point cloud data of the chip.(2)The point cloud data of chip pre-processing include coordinate correction,filtering,simplification and so on.The coordinate correction is to correct the X coordinates of the point cloud,according to the optical path structure of the line laser sensor.A density based method is used to remove outliers in point cloud,and bilateral filtering algorithm is used to smooth the point cloud surface and eliminate irregular points on the surface.The point cloud reduction based on voxel raster sampling and the point cloud reduction based on the curvature of point cloud are compared.Finally,according to the actual situation of the point cloud,the point cloud data are reduced by using the voxel grid sampling method.(3)Dividing the chip point cloud and locating the pin of the chip.In order to extract the pin part separately from the point cloud,using random sampling conformance algorithm(RANSAC)for plane segmentation.After fully analyzing the advantages and disadvantages of the K mean clustering algorithm,An improved K mean clustering algorithm is used to locate each pin of the chip.(4)In order to evaluate the coplanarity of chip pins,the least squares method is used to fit the plane of each pin of the chip,and then calculate the distance from each pin to the fitting plane.The experimental results show that the method used in this dissertation can be used to detect the coplanarity of the chip very well,and meet the non-contact and high precision detection requirements.