X-RAY即時影像強化與瑕疵偵測之研究

Real-time X-ray Image Enhancement and Defects Inspection

李淑惠(Shu-Hui Lee)

指導教授 ： 江行全

元智大學/工程學院/工業工程與管理學系/碩士(2001年)

摘要

X射線檢測法由於可以觀察物體內部結構，因此應用領域廣泛，常應用於工業或醫療領域。由於X射線與物質作用時會產生穿透、散射、吸收現象，因此，待測物的密度與X射線照射的強度是影響影像品質的重要因子。所以當產品組成零件之材料越複雜，X-ray影像的品質也越難控制。
本研究所檢測之國防工業產品有下列兩項特性：
1.物件組合材料複雜，使得所擷取的影像模糊。
2.物件為一旋轉物件。
針對這些問題，本研究分為兩個部分，一為影像即時強化，二為影像比對。影像強化階段，本研究針對目前線上不同的兩種產品，根據產品影像特性，建議兩種不同的強化方式。比對階段，使用相關係數法進行樣板比對偵測出目標物，並對目標物進行相關係數計算，判斷產品是否有異常的發生。
實驗結果發現，強化階段，可以達到即時強化的效能，並協助線上操作人員檢測。比對階段，相關係數法大致上可以正確地找出目標物，但是目標物傾斜則容易產生誤判。

關鍵字

影像強化 ； 樣板比對 ； 相關係數 ； X-ray

並列摘要

X-ray inspection has wide applications in industry, because it can observe the structure inside of an object. X-ray has the characteristics of perspective, scattering, and absorption through the object, so there are important factors that the object density and the X-ray intensity have effects on image quality. If a product consists variety of materials, it would be more uncontrollable of image quality in using X-ray. The focus of this study is an application of defence industry products which have two characteristics. One of the characteristics is to be complex of the products materials, and the other is to be rotatable of the object.
According to these problems, this research focuses on tow subjects: one is real-time X-ray image enhancement, the other is image matching. The first subject of image enhancement, two different methods based on image feature of two products on the production line are recommended. The second is using correlation method of template matching method to find the target and detect the target which is good or not.
The experiment results of image enhancement stage show that the research improve the original X-ray image quality in real-time process by using digital image processing, and assist the operators in inspecting products. Then the results of the image matching stage can almost detect the defects certainly. In this study, the matching capability is 95%.

並列關鍵字

X- Ray inspection ； image enhancement ； template matching ； correlation

參考文獻

Boerner, H., H. Strecker, “Automated X-ray inspection of aluminum castings,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 10, No. 1, 1998, pp. 79-91.

Hansen, K. A., “Nondestructive testing of wooden vessels,” Proceedings of the MTS/IEEE Conference, Vol. 2 , 1997 , pp.1173—1187.

Murray, N.C., R.J. Lacey, and P.H. Mason, “ Exploitation of X-ray technology for the detection of contraband-aviation security applications,”Proceedings of the European Conference on Security and Detection , 1997 , pp. 13 —18.

Smith, G.J. “Bodysearch technology uses X-ray imaging to remove hazards and humiliation from personnel searches,” Proceedings of the IEEE 29th Annual 1995 International Carnahan Conference on Security Technology, 1995 , pp.502 —504.

Avi-Itzhak, H.I., J.A. Van Mieghem, L. Rub, ”Multiple subclass pattern recognition: A maximin correlation approach,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 17, No. 4, 1995, pp. 418-431.