立體視覺於工具機加工碰撞與原點偵測技術之探討

Assessment of Practicality of Collision Prevention and Work Offset for Tool Machine Using Stereo Vision

簡睿弘(Jui-Hung Chien)

指導教授 ： 江佩如

國立中正大學/工學院/機械工程學系/碩士(2016年)

摘要

工具機的操作者必須要耗費心力確認工具機的環境與監控工具機的運行，當操作者的技術不熟練時，更是容易發生錯誤，由於工具機並無一套系統可以確認實際環境的狀況給予操作者客觀的資訊，使得使用工具機的技術門檻變高，人員的培訓難度增加，因此本研究利用立體視覺系統輔助工具機的使用，重建刀具位置與工件邊緣，並計算兩者之間三軸向的距離，最後進行碰撞與加工原點偵測，碰撞偵測是在刀具向工件的移動過程中，量測刀具與工件邊緣的最近距離，得到的最小誤差為0.130mm，最大誤差為8.440mm，若偵測距離小於15mm則判斷為危險，本研究在8組不同距離下進行碰撞偵測，結果皆為正確。加工原點偵測為當機台回歸至機械原點時，量測加工原點(Part Zero)與機械原點(Machine Zero)的距離，X軸、Y軸與Z軸的平均誤差分別為0.751mm、2.928mm、1.538mm，雖然不夠準確，但是提供操作者更為客觀的距離判斷。

關鍵字

立體視覺 ； 尋邊 ； 工具機座標 ； 影像處理

並列摘要

Machine operation training is a long term work. To ensure the safety of the operators and tool machines, the condition of the tool machine has to be checked once and again before and during the operation. The machine is especially easy to break down under unskilled operation. Because of the lack of automatic condition surveillance, operators have to be experienced and pay attention during the operation. Thus, this paper proposes a stereo vision system to help operators check the condition of tool machine. This system mainly measures the distance between the tool and edges of the work piece such that the work offset can be done automatically and collision can be prevented. In the collision detection task, distances of different locations of the tool and the work piece were measured. The minimun error is 0.130 mm and the maxmum error is 8.440 mm. When the detected distance was larger than 15mm, the tool condition was judged as safe status, otherwise collision warning was issued. Among 8 measurements, all judgements are correct. In the task of work offset, when the machine returns to machine zero, the distance between part zero and machine zero were measured. The average error of X-axis、Y-axis and Z-axis are 0.751mm, 2.928mm and 1.538mm, respectively. In spite of the lack of precision, this system helps operators to understand the conditions of the tool machines.

並列關鍵字

Stereo vision ； Work offset ； Tool machine coordinate system ； Image processing

參考文獻

[1] X. Tian, X. Zhang, K. Yamazaki, and A. Hansel, "A study on three-dimensional vision system for machining setup verification," Robotics and Computer-Integrated Manufacturing, vol. 26, pp. 46-55, 2010.

[2] R. Ahmad, S. Tichadou, and J.-Y. Hascoet, "New computer vision based Snakes and Ladders algorithm for the safe trajectory of two axis CNC machines," Computer-Aided Design, vol. 44, pp. 355-366, 2012.

[3] X. Zhang, W. Tsang, X. Tian, K. Yamazaki, and M. Mori, "Automatic segmentation of the apparent contour for 3D modeling of cutting tools from single view," in Advances in Visual Computing, ed: Springer, 2008, pp. 772-781.

[4] X. Zhang, W.-M. Tsang, K. Yamazaki, and M. Mori, "A study on automatic on-machine inspection system for 3D modeling and measurement of cutting tools," Journal of Intelligent Manufacturing, vol. 24, pp. 71-86, 2013.

[5] 佘東穎, "應用立體視覺3D量測技術於五軸工具機防碰撞," 碩士論文, 電機工程研究所, 國立中正大學, 2014.