影像處理與視覺技術於機械金屬領域之應用

Application of Image Processing and Machine Vision Technology in Mechanical and Metallic Fields

許光城(Quang-Cherng Hsu) ； 梁軒齊(Xuan-Qi Liang)

《鍛造》 28卷3期 (2019/10) Pp. 17-26

摘要

工業4.0智慧製造時代來臨，機械金屬領域亦不能自免於此潮流，而智慧製造需要搜集大量資料，感測器便是取得即時資料的重要來源，其中影像感測器具有類似人眼取像功能，若再搭配電腦分析、辨視、量測、檢測等模型，則有機會達到智慧製造的境界。本文先簡述影像處理與機械視覺技術的理論背景，並針對機械金屬領域，例如：板金成形量測（2D應變、3D應變、異向性量測）、鍛造品3D量測提昇材料使用率、金屬製品量測（車用不等徑彈簧）、機械製品量測（類比式儀錶指針數字錶號、車用單向軸承檢測、玻璃基板邊緣缺陷檢測）、塑膠製品量測（No-kink水管彈回量測、牙線棒半成品瑕疵檢驗），最後以工業機械人與電腦視覺之應用作為總結說明影像處理與視覺技術於機械金屬領域之應用。

並列摘要

Smart manufacturing such as Industry 4.0 is coming, the mechanical and metallic fields can't avoid this trend, and however, it needs to collect a lot of data. The sensor is an important source of real-time data. The image sensor has similar human eye imaging function. If you use models such as computer analysis, recognition, measurement, and detection, you will have the opportunity to achieve the realm of smart manufacturing. This paper first briefly describes the theoretical background of image processing and machine vision technology, and for the field of mechanical metals, such as sheet metal forming measurement (2D strain, 3D strain, anisotropy measurement), forging 3D measurement to improve material utilization rate, Metal product measurement (vehicle unequal-diameter spring), mechanical product measurement (analog meter pointer digital meter number, vehicle single direction bearing inspection, glass substrate edge defect detection), plastic product measurement (No-kink water pipe spring-back measurement, dental floss pick inspection), and finally the application of industrial robots and computer vision as a summary of the application of image processing and vision technology in mechanical metallic fields.