視覺檢測應用於水平多關節機械手臂取料系統

Using of Visual Inspection for SCARA Robotic Arm Pick and Place System

蔡詠任(Tsai Young-Ren)

指導教授 ： 林宜弘

國立屏東科技大學/工學院/機械工程系所/碩士(2022年)

https://doi.org/10.6346/NPUST202200193

摘要

本論文以影像辨識技術為基礎，對料盤中之特定範圍的目標物件進行辨識與分類，並傳送相關資訊給水平多關節機械手臂，使手臂依所設定之路徑運作，完成辨識、取料、放料等一系列流程。
本系統包含影像辨識與機械手臂，在影像部分使用攝影機配合光源模組，對料盤中紅、綠、藍與黃四種顏色的長方形料件進行影像擷取，並透過RGB通道抽取與灰階辨識各個物件的顏色，接著藉由EMGUCV函式庫繪製出物件的位置與大小，再利用圖像矩計算物件的角度；而機械手臂則負責接收物件的顏色、位置與角度等資訊，吸取對應座標的物件並進行角度轉正並依據物件顏色與所設定之放料區座標進行放料。
在獲得物件座標後，因為影像座標單位與機台座標單位不同，需要先進行單位換算，將影像單位(像素)換算為機台單位(毫米)後，再透過座標系校正將影像座標系轉換為世界座標系。

關鍵字

影像辨識 ； 彩色影像 ； 水平關節機械手臂 ； 取料與放料

並列摘要

Based on image recognition technology, this paper identifies and classifies target objects in a specific area of the pallet and transmits relevant information to the SCARA robot arm, which operates according to a set path to complete a series of processes such as identification and pick-and-place.
The system architecture consists of image and robot arm. In the image part, a CCD camera with optical module is used to capture images of rectangular objects in red, green, blue and yellow colors in the charging tray, and identify the color of each object through RGB channel extraction and grayscale, then draw the position and size of the object and calculate the angle of the object by using the EMGUCV library. The robot arm is responsible for receiving the color, position and angle information of the object, picking up the object with the corresponding coordinates, making angle rotation and discharging according to the object color and the set coordinates of the discharge area.
After obtaining the object coordinates, because the image coordinate units are different from the machine coordinate units, it is necessary to convert the image units (pixels) to machine units (millimeters) and then convert the image coordinate system to the world coordinate system by coordinates system correction.

並列關鍵字

Image Recognition ； Color Image ； SCARA Robotic Arm ； Pick and Place

參考文獻

1. Selig, J. M., 1992, “Introductory Robotics”, Saddle River, Prentice Hall.

2. Maneetham, D., 2018, “Kinematics Modeling and Simulation of SCARA Robot Arm,” Mechatronics Engineering, Rajamangala University of Technology Thunyaburi , Thailand.

3. Taylan Das, M. and Canan Dulger, L., 2005, “Mathematical modelling, simulation and experimental verification of a scara robot,” Department of Mechanical Engineering, University of Gaziantep, 27310 Gaziantep, Turkey.

4. Ahn, I., Zo, M. M. and Kim, C., 2009, “A robust bread defect detection and counting system,” International Conference on Advanced Communication Technology, Paper No. 10571395.

5. 唐滄澤，2010，互動式影像操控-基於手勢及顏色辨識與控制，碩士論文，國立中正大學，電機工程所，嘉義。