摘要
模板匹配是數字圖像處理中的重要內容，它的作用是在原圖像中找到于模板圖像最相近的部分。
本文對模板匹配進行學習和研究，主要介紹了一種快速的匹配算法——SSDA(Sequential Similarity Detection Algorithms)算法。SSDA算法是將模板在原圖上平移比較，平移到最后找出相似度最大的一個部位，就是我們需要找的。在SSDA算法中用了三個不同形式的公式進行編程，通過對不同圖像的處理，各個公式對圖像處理的時間和精確度進行比較。
在Microsoft Windows XP平臺上Visual C++ 6.0開發環境下，采用上述方法編制了模板匹配的軟件。實驗證明，該軟件對于尺寸不大的清晰圖片基本能夠正確定位，并進行識別，典型耗時為0～3秒鐘。
最后，本文探討了SSDA算法的方法的優缺點，提出了進一步工作的方向。
關鍵字：圖像匹配 序列相似性 SSDA算法 圖像模板 VC++
Abstract
The template match is an important content in the digital picture processing; its function lies in finding the part which is closest to the template picture from the original map.
The template match has been studied and researched in this paper, and the emphasis has been put on the SSDA (Sequential Similarity Detection Algorithms) algorithm, which is one kind of fast match algorithm. The SSDA algorithm is to translate and compare the template in the original map, to finally discover the spot which has most greatly similar. Three different forms formulas in the SSDA algorithm have been carried on the programming, and each formula carried on the comparison to the picture processing time and the precision.
In the platform of Microsoft Windows XP and the development environment of Visual C++ 6.0, the software about the template match has been compiled using the above method. The experiment proved that, this software can correctly locate and recognize these kind of pictures which are not big in size but clear with the model time consuming at 0 ~3 seconds in usual.
Finally, the good and bad points of the SSDA algorithm method have been discussed, and the further direction of research has been proposed in this paper.
Key words: Picture match Similar of sequence SSDA algorithm Picture template VC++
