面向视觉伺服的芯片图像Zynq预处理系统  

Chip Image Preprocessing System with Zynq for Visual Servo

作　　者：景鹏阁 龚时华[1] 周丹宇 于思恂 JING Pengge;GONG Shihua;ZHOU Danyu;YU Sixun(College of Mechanical Science and Engineering,Huazhong University of Science and Technology,Wuhan 430070,China)

机构地区：[1]华中科技大学机械科学与工程学院,武汉430070

出　　处：《计算机测量与控制》2024年第11期287-293,共7页Computer Measurement &Control

基　　金：国家自然科学基金(52175462)。

摘　　要：晶圆芯片定位精度直接影响到芯片制造的效率和质量,视觉伺服作为一种实时控制方法,引入了视觉反馈,不仅能够保证芯片定位精度还具有良好的鲁棒性;同时图像处理的实时性对视觉伺服至关重要;传统的使用PC机和网线的图像处理方式实时性和准确性较差,为了提高这一过程中视觉反馈的精度和实时性,提出并设计了一套面向视觉伺服的芯片图像Zynq预处理系统;系统通过在Zynq PL端的片内设计和PS端的控制设计,对采集到的芯片图像进行去噪、Sobel处理,并给芯片粗定位,为芯片的伺服控制提供有效数据;经实验表明,该系统实时性强,位置识别稳定,达到设计要求。The positioning accuracy of wafer chip directly affects the efficiency and quality of chip manufacturing.As a real-time control method,visual servo introduces visual feedback,which not only ensures the positioning accuracy of chip but also has a good robustness.At the same time,the real-time performance of image processing is crucial for visual servo.Traditional image processing methods for using PC and network cables have the characteristics of poor in real-time and accuracy.In order to improve the accuracy and real-time performance of visual feedback in this process,a Zynq preprocessing system for visual servo-oriented chip images is proposed and designed.Through the on-chip design of Zynq PL terminal and the control design of PS terminal,the system performs the denoising,Sobel processing on the acquired chip images,and rough positioning of the chip,which provides effective data for the servo control of the chip.Experimental results show that this system has a strong real-time performance and stable location recognition,meeting the design requirements.

关 键 词：芯片制造 Zynq FPGA 视觉伺服 图像处理 

分 类 号：TN911.73[电子电信—通信与信息系统]