在早期可用來顯示的裝置並不多，不外乎就是電腦monitor及電視…等，而這些顯示裝置的輸入界面大概就幾種而已，如D_SUB in及 AV in 及S端子in…等。為使顯示裝置能正常顯示，故在生產這類有輸出顯示界面產品時，就必須針對這些信號做檢測；在以往較少樣的顯示界面檢測方式較單純，且生產流程上也不需要特別添購許多設備，故以人眼去檢測似乎是比較可行且C/P值也較高。 但隨著科技日新月異且消費者對影像及聲音的品質愈來愈要求下，相對的考驗生產廠商的檢測能力，且現今消費者信手拈來即是智慧型手機再加上網路普及化，若某產品有重大的功能瑕疵，短時間就會在網路上散佈開來，那這家廠商生產的所有產品，就有可能會被消費者列入不考慮購買之廠商。 故為了解決這個問題，勢必要找出一個可行且排除以人眼來判定這些顯示界面的好壞。 本論文即以此為出發點做探討，將這些顯示界面信號經過擷取裝置到電腦後做影像分析，並搭配標準測試檔來達到以下檢測目的： (1) 優化色彩檢測方法。 (2) 排除人眼主觀判定影像正確與否。 (3) 在生產流程上可做到一人多機測試。 當然要建立這些標準檔前，必須先了解影像分析的基本理論後再進行程式開發，方能將理論與實務做結合。 ;There are not many kinds of interface cards for displaying color signals used in TVs or computer monitors. To comply with these interfaces of display device, the signal testing is required in the production of such output interface products. In the past, the testing of these kinds of interfaces was simple and needs equipments of a few instruments. Usually eye detection can satisfy this purpose. Because modern high quality video signal interface cards may require much advanced techniques, the testing departments need to combine variety kinds of interface cards. Therefore, the eye testing method cannot meet this requirement. This thesis first analyzes the displaying signal sent from the image capturing device. Then it optimizes the testing process through the following functions: 1. Standardizing and unifying the testing process. 2. Excluding the factor of eye detection. 3. Handling multi-machine in the production process. A real testing system is established to test this optimization for displaying color signal and sound interface cards. We utilized the software named as Direct Show and Graph Edit function by using Microsoft Visual Studio 2010.The results much improved the optimization of the testing procedures.
