Research On Wafer Chip Recognition And Location Algorithm For Visual Servo

With the electronic manufacturing equipment industry continues to develop,electronic manufacturing equipment efficiency and accuracy requirements continue to improve.In the field of chip solid crystal,detection,sorting and other fields,the need for the movement platform can quickly and accurately identify and locate the wafer chip,positioning speed and accuracy directly determine the efficiency and quality of electronic manufacturing.In this paper,we use machine vision and image processing technology to study the chip recognition and location algorithm for visual servo.The main work includes:(1)In order to meet the requirements of high speed and high precision positioning of visual servo platform,firstly the article analysis the function requirement,and the time series analysis of machine vision system.Real-time development of machine vision system is realized by real-time extension software KRTS.Then the hardware is determined for the light source,camera and motion control system,and a visual servo control model based on machine vision and motion control is constructed.(2)In order to improve the accuracy of recognition and location,the edge region of the chip is first coarse-coded,and then the article use the improved adaptive operator to remove the edge interference.Finally,the article use the least squares fitting to remove the large error point,effectively improve the accuracy.(3)In order to improve the speed of chip identification and positioning,we first optimize the search area and search angle,and then propose an improved shape-matching algorithm,and introduce three strategies to reduce the computational complexity and improve the matching speed.(4)This paper adopts VS2013 developed a wafer chip identification software,realize the image acquisition and display,wafer chip identification and positioning,and carries on the experiment in view of the actual chip.The chip edge-detection algorithm focused on the sub-pixel level Precision performance,the improved shape-matching algorithm focused on the rapid identification of its performance.The experimental results show that the proposed algorithm and its processing speed and accuracy meet the actual requirements of production.