背光滑鼠瑕疵機器視覺檢測系統

Backlit Mouse Defect Inspection Using Machine Vision

陳柏安(Po-An Chen)

指導教授 ： 彭德保

國立交通大學/管理學院/工業工程與管理系所/碩士(2012年)

https://doi.org/10.6842/NCTU.2012.00315

摘要

近年來隨著個人電腦的普及，電腦周邊產品也隨之蓬勃發展，製造商為了加強競爭優勢，因此致力於研發以消費者市場為導向之新一代電腦周邊產品。以普遍使用的電腦周邊產品滑鼠為例，新一代的滑鼠擁有背光圖騰設計，其背光區塊可能於製程中不可預期的因素，產生瑕疵。目前業界以人工方式進行背光區塊之圖騰檢驗，耗時、成本高、檢驗成效又不穩定；有鑑於此，本研究針對其常見之
(1)缺角、(2)裂邊、(3)斷裂，以及(4)紅色灰階值異常等四項瑕疵，開發出快速又正確穩定之瑕疵檢測系統。檢測系統分為訓練階段與檢驗階段；在訓練階段中，以一批標準的滑鼠影像，運用影像處理方法，萃取其相關特徵值，並以統計分析方法，利用管制上下界定義出對應參數規範。而後，在檢驗階段中，以離線單機作業的方式，針對待測之背光滑鼠影像，經由影像前處理、影像切割與特徵擷取後，參照訓練階段所得之規範，自動判斷滑鼠是否有瑕疵。本論文所提之背光滑鼠瑕疵機器視覺檢測系統，能快速又正確的進行自動檢測，可用來取代傳統人工檢測，以防杜不良品外賣，提升產品品質與品牌形象。

關鍵字

背光滑鼠 ； 機器視覺 ； 自動光學檢測 ； 瑕疵檢測

並列摘要

Mouse is one of computer peripheral products that have widely been used. To enhance the competitive advantage, manufacturers need to develop the consum-er-oriented mouse. A new generation of mouse with the backlit patterns has been is-sued to attract the gamers. However, the backlit mouse might have the defect after producing or assembly. The common defects include (1) incorrect illuminating area, (2) crack, (3) fragment, and (4) incorrect LED color saturation. Traditionally, the manufacturers still use human inspection which might be time-consuming, high labor cost, and unstable testing results.
This research is to develop an auto-inspection system which can inspect the mentioned defect of the backlit mouse and replace the human inspection. The pro-posed system consists of a training phase and a testing phase. The training phase is to train the upper control limit (UCL) and the lower control limit (LCL) of each corresponding defect type from a number of standard images. Then the defect fea-tures of the to-be-inspected images was extracted and checked whether by the proposed method in the testing phase. Experimentation results revealed that the proposed system can inspect the four defective types for the backlit mouse effectively and robustly.

並列關鍵字

backlit mouse ； machine vision ； automated optical inspection ； defect inspection

參考文獻

【1】 H. Freeman, Machine Vision: Algorithms, Architectures, and Systems, Aca-demic Press, 1988.

【6】 P. K. Sahoo, S. Soltani, and A. K. C. Wong, “A survey of thresholding tech-niques,” Computer Vision, Graphics, and Image Processing, vol. 41, pp. 233-260, 1988.

【8】 P. Soille, Morphological Image Analysis: Principles and Applications, 2nd ed., Springer, 2004.

【9】 E. R. Davies, Machine vision: theory, algorithms, practicalities, 2nd ed.: Academic, 1997.

【10】 R. C. Gonzalez and R. E. Woods, Digital Image Processing, 2nd ed., Pren-tice Hall, 2002.