基於PLC與PC-based軸控之雙相機系統自動化排列機開發

機器視覺檢測系統通常只使用一組工業攝影機，但如果檢測的樣品同時具備數量大、體積小、多種形狀和顏色的條件，則單一組工業攝影機所能夠拍攝的影像範圍和解析度是有限的。本研究提出一雙相機視覺檢測系統，架設兩組影像解析度都為0.162 mm的工業攝影機，分別拍攝一個面積為27.5×37 cm2檢測盤面的兩個區塊，檢測的樣品為4種形狀、6種顏色，總共24種面積都不大於1 cm2的壓克力，再搭配自行開發的自動化排列機，吸取樣品後進行樣品排列。本研究依據24種樣品不同的顏色、形狀及灰階值，使用影像學習建立資影像資料庫，並在第一次啟動人機介面時執行，執行時間約為22.434秒。影像辨識則根據影像資料庫，將檢測的樣品定義為與數值最相近的種類，一個樣品的辨識時間約為0.024秒。龍門自動化排列機之X、Y軸的作動元件為導螺桿，Z軸為氣壓缸，A軸則為步進馬達。樣品的抓取是利用真空吸取辨識後的樣品，旋轉A軸以改變其放置角，再移動到指定位置。另外，本研究亦分別使用PLC與PC軸卡控制機台，比較兩者對排列速度的影響，結果PLC的平均單顆速度約為1.737秒，PC-based約為1.853秒。

關鍵字

機器視覺；雙相機；自動排列；影像辨識； PLC與PC-based差異

並列摘要

Machine vision inspection system is only taken an industrial camera typically. But if the inspected samples are accompanied by a large quantity, a small size, a variety of shapes and colors. The scope and resolution of the image is limited when using only one industrial camera. This paper developed a dual cameras inspection system, set up two industrial cameras and image resolutions are 0.162 mm. They shoot two parts of the inspection board with an area size of 27.5×37 cm2. The samples to be inspected have four shapes and six colors, the thickness is 1 mm and the area is no more than 1 cm2 of acrylic. Using the self-developed gantry automatic sorting machine to suck up the samples and sort them. In this study, the image learning is based on twenty-four samples of colors, shapes and grayscale values to establish the database, and performs that when the HMI starts for the first time. The image learning takes about 22.434 sec. The image recognition is based on the image of the database, and the detection of the sample is defined as the most similar with the type of value. The image recognition takes of a sample is about 0.024 sec. The gantry automation sorting machine of the X, Y-axis actuating element is the lead screw, the Z-axis is the pneumatic cylinder, and the A-axis is the stepping motor. The machine uses a vacuum to suck up the recognized sample, rotate the A-axis to change its placement angle and move to the specified position. In addition, this study also using the PLC and PC axis card control the machine to compare the sorting speed. The research shows that average single time of the PLC and PC-based are respectively 1.737 sec and 1.853 sec.

並列關鍵字

Machine vision ； dual cameras ； automatic sorting ； image resolution ； PLC and PC-based difference

參考文獻

[1] Pu Huang, Zhangjing Yang, Caikou Chen, 2015, “Fuzzy local discriminant embedding for image feature extraction”, Computers and Electrical Engineering, 46, pp.231-240.

[2] Emilia López-Iñesta, Francisco Grimaldo, Miguel Arevalillo-Herráez, 2016, “Combining feature extraction and expansion to improve classification based similarity learning”, Pattern Recognition Letters, 000, pp. 1-9.

[3] Chandra Sekhar Nandi, Bipan Tudu, Chiranjib Koley, 2014, “A Machine Vision-Based Maturity Prediction System for Sorting of Harvested Mangoes”, IEEE Transactions on Instrumentation and Measurement, 63(7), pp. 1722-1730.

[5] Sofu, M. M., O. Erb, M. C. Kayacan, B. Cetişli, 2012, “Design of an automatic apple sorting system using machine vision”, Computers and Electronics in Agriculture, 127, pp. 395-405.

[6] Jyoti Jhawar, 2016, “Orange Sorting by Applying Pattern Recognition on Colour Image”, Procedia Computer Science, 78, pp. 691-697.

國際替代計量

基於PLC與PC-based軸控之雙相機系統自動化排列機開發

未授權

主題瀏覽