基於影像伺服控制之XXY定位平台的設計與實現

Design and Implementation of XXYPositioning Stage based on Vision Servo Control

廖偉豪(Wei-Hao Liao)

指導教授 ： 林志哲

國立臺北科技大學/機電學院/自動化科技研究所/碩士(2012年)

摘要

本研究的目標主要是以XXY平台來影像伺服控制，建構一套自動化對位系統，來取代一般人工對位，結合影像視覺技術與XXY三軸之微步進平台來開發一套可用於XXY平台上的影像伺服自動化控制系統。
研究中以壓電噴墨頭的組裝試片作為系統之對位組裝測試之元件，並透過C++軟體、GIGAE CCD與鏡頭、以及DSP運動控制卡來整合XYY三軸定位平台系統與影像辨識系統二部份，完成精密的影像伺服定位。影像辨識系統來針對待測物上之對位符號偵測，以Sobel運算找出對位圓之輪廓，在以最小平均方誤差法來求其半徑與圓心位置，另外配合對位符號以霍式轉換的方式來對直線作偵測。為了使定位組裝的精度更為精確，本研究應用類神經網路系統來進行平台因機構本身因素所造成的位移誤差之校正，可以提高應用影像辨識之伺服定位精度。
本研究中實驗主要探討XXY平台可直接在平台上任一點進行原地旋轉，並模擬上下試板對位時的情形，結合影像與平台控制技術來完成自動化對位。

關鍵字

XXY定位平台 ； 影像伺服控制 ； 十字標靶

並列摘要

In this study, the micro-assembling task of the positioning system is discussed; an automatic micro-assembling system is built to replace manual operations by XXY three-axis micro-stepping positioning stage with the image recognition system.Study of a piezoelectric inkjet head assembly of the specimen as assembly and testing of system components and C++ software, GIGAE CCD and the lens, DSP motion control card to integrate the XYY three-axis positioning stage system with image recognition system part of the completion of the sophisticated vision-servo positioning.For the coarse positioning task, the Sobel operator is used to find outline of positioning mark and we used Least-Mean-Square Error Algorithm to calculate center and radius. For the fine positioning task, the Hough transform is applied to find the straight line of the cross positioning mark.To result of this study show that, we use the neural network system improve to the stage of the institutions themselves caused the displacement error correction. We can improve the servo positioning accuracy of the application of image recognition.

並列關鍵字

XXY Micro-positioning Stage ； Vision Servo Control ； Line Detection

參考文獻

[18] 全研科技股份有限公司，四驅動對位對準平台，智慧財產局專利。

[19] 全研科技股份有限公司，XXY平台型錄規格表，彰化，台灣全研科技股份有限公司。

[1] Y. Hara, N. Akiyama, and K. Karasaki, Automatic inspection system for printed circuit boards, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. PAMI-5,NO.6, Nov. (1983) 41-49.

[3] J.M. Parker, A robust machine vision design to facilitate the automation ofsurface appearance inspection, Proceedings of IEEE/ASME InternationalConference on Advanced Intelligent Mechatronics, July (2001) 88-92.

[4] Z. Wei, G. Zhang, and X. Li, The Application of Machine Vision in InspectingPosition-Control Accuracy of Motor Control System, Proceedings of the Fifth