機器視覺技術已被廣泛地應用在各個領域，例如製造業生產線的瑕疵偵測技術。傳統的機器視覺系統開發流程，需耗費大量的開發時間與成本。本論文基於Grafcet描述語言的特性，設計具有快速開發與跨平台能力的機器視覺系統開發平台。此一平台使用開放源碼的Beremiz編輯器編輯Grafcet，透過平台解析Grafcet XML檔案中的流程步驟，取得各步驟的功能模組名稱與參數後，經過相關的影像處理流程後，呈現執行結果。結合此兩種軟體，搭配成為一個具有快速開發、驗證能力的開發平台。基於Grafcet描述語言，將嵌入式系統的架構與各功能模組以Grafcet描述語言建立離散事件模型，在離散事件模型中可以清楚地知道所有功能模組的動作、控制條件與狀態轉移。此一方式設計的主要重點為提供一友善的使用者介面與清楚明確的功能接口函數，讓系統使用者對所需的功能採用模組化的配置方式開發，可以縮短開發時間與成本，即時觀看執行結果。;Machine vision technology has been widely applied in various fields, such as manufacturing line defect detection. The traditional processes to development machine vision take a lot of time and cost. In this paper, we proposed a development platform, based on Grafcet description language features, with rapid and Cross-Platform abilities to design machine vision systems. We use an open source software Beremiz as Grafcet editor and parse the process steps, function names and parameters from Grafcet XML file through the platform. It’s capable of rapid development and validation by use of this methodology. Based on Grafcet description language to create discrete events for the embedded system can clearly show all the actions, conditions and state transitions of the architecture. The advantage of this approach is to provide a user-friendly and clear functional interface for users modular-configuring development. It can effectively improve the system design tasks and complexities.
