隨著物聯網 ( Internet of Things ; IoT ) 的趨勢發展，工業應用領域也開始推動工業4.0的各種整合應用。所謂工業4.0概念，最早提出在2011年的漢諾威工業博覽會，由德國政府提出的高科技戰略計畫，其概念將傳統製造業運用IOT的架構，得以轉型成具有高度適應性、高資源效率，進而發展成全面自動化生產的智慧工廠。 過去傳統產業所強調的創新，即為創造差異化，專注在成本、品質、產能、交期等目標上，但現在趨勢是以客戶為導向，考驗的是企業的應變彈性，必須在客戶有需求時調整製程、快速供貨。 因此企業的製程需要具備良好的應變能力，妥善運用數據資料的蒐集、儲存管理、分析應用。以大數據技術，善用收集數據並預測模式導向；發展出最合適又具效益的新商業模式，逐步邁向生產自動化、智慧化工廠的未來。 在本論文中，選定電池的品質檢測為研究對象，在現有的人工目視檢測製程中，因人員疲勞以及不同人員判定標準不一致，難以建立品質管理之標準。 因此本研究提出一套機器視覺檢測技術來取代現有的人工目視檢測方式，針對電池邊緣與表面的瑕疵檢測技術，進行分析與研究。經本論文實證，此技術可確實將瑕疵數據化，在此基礎上得以建立品質管理的標準，並將檢測速度由現有的每秒1顆提升至每秒5顆以上，將此檢測製程藉由此機器視覺檢測技術來達成高精度、快速、低成本的生產製程，得以提升產品良率並且數據化品質管控的標準。;With the development of the Internet of Things, industrial applications start to promote integrated applications of Industry 4.0. The Industrial 4.0 concept was first proposed at the 2011 Hannover Industrial Fair by the German government. This concept is to use IOT technology and help the traditional manufacturing transformed into a smart factory. In the past, traditional manufacturing focus on innovation, which was to create differentiation and focus on the goals of cost, quality, production capacity, delivery, etc. However, the market now is customized products. The delivery process must be quickly adjusted and customized. Therefore, the manufacturing needs to have a good ability to adapt by use the data collection and analysis. Based on the Big Data application, it will develop a new business model and transform into a smart factory. The thesis is mainly concerned with the surface defect issue of the battery. It proposed the automatic optical inspection system to replace the manual detection and also provides high-precision, fast, and low-cost production processes. Not only improve product yield rate and digitize the quality control standards
