- 回首頁
- 研究與發展
研究與發展
導磁材料之AI智慧磁檢探傷模組
2023/07/13
簡介 Introductions
- 本模組結合傳統機器視覺及深度學習之優勢並整合螢光顯示技術,以期解決現有針對表面裂痕其螢光顯影後磁檢設備商無法智慧檢測問題,透過此導磁材料之AI智慧磁檢探傷模組,進行可疑瑕疵檢測複判機制,減少人員複檢負擔並可上傳雲端更新檢測模型,增加螢光磁檢智慧探傷功能、提升國內磁檢設備競爭力。
- This module combines the advantages of traditional machine vision and deep learning, and integrates fluorescent display technology, to intelligently inspect surface cracks after magnetic testing, a problem that magnetic particle inspection (MPI) equipment manufacturers are unable to do. Through the AI smart magnetic particle inspection module of ferromagnetic materials, a re-judgment mechanism for suspicious defect inspection can be carried out. This reduces the burden of the operator and by uploading updates of the inspection model to the cloud, it increases the intelligent function of magnetic particle inspection; thereby, enhancing competitiveness of domestic MPI equipment.
特色與創新 Characters and Innovations
- 藉由AI技術導入智慧檢測,克服傳統AOI設備,常因螢光磁粉殘留造成誤判率 > 20%,轉換為具有人工智慧檢測能力之螢光磁檢設備,達成自動智慧探傷之目的。
- 透過可疑瑕疵檢測複判機制,減少人員複檢負擔並可上傳雲端更新檢測模型。
- Intelligent inspection through AI technology to overcome issues in traditional AOI equipment, which often has a misjudgment rate >20% due to residual fluorescent magnetic powder. Converts magnetic particle inspection equipment with artificial intelligence inspection capabilities to achieve automatic intelligent MPI.
- Through the suspicious defect re-judgment mechanism, the burden of the operator to re-check can be reduced and the inspection model can be uploaded to the cloud for updating models.
應用與效益 Applications and Benefits
- 可降低75%人力需求,將人員素質提升至做更高等級工作,另改為AI檢測設備檢測可全天24小時生產並檢測,加值設備價值至少提升兩倍以上,大幅增加設備效能與競爭力。
- This module can reduce manpower requirement by 75%, and improve the quality of personnel to do higher-level work. The AI inspection equipment can detect and produce 24 hours a day, which doubles the value of the equipment; thereby, greatly increasing equipment efficiency and competitiveness.
聯絡 contact
AI智慧磁檢探傷模組 AI Smart MPI Module
機邊複判與雲端模型主動更新機制 Edge Defect Re-check and Cloud Model Self-Updating Mechanism