由於現今資訊科技的蓬勃發展，人們運用各種方法針對例如指紋等生物特徵，以及例如刺青等軟性生物特徵進行檢索，然而，在海量般的巨大影像資料庫中，如果只運用指紋等單一生物特徵搜尋資料，將花費大量之時間與資源，如果能輔以如刺青等軟性生物特徵來進行鑑定識別的動作，對檢索效能將提供很大的幫助。另外，資訊檢索技術提供了人們在進行犯罪現場辨識、嫌疑犯辨識，受害者身分鑑定等資訊鑑識活動中，大幅提昇了辨識資訊的便利性。
本論文運用內容式影像檢索廣泛使用之顏色空間直方特徵，完成一套刺青影像檢索系統。系統首先針對刺青影像資料庫進行最小方形之擷取，再對每一刺青影像加以分割為九等分，紀錄其顏色空間特徵，再配合對刺青影像資料庫之色彩、類別加以分類，最後以查詢之刺青影像，對刺青影像資料庫進行相似度比對。實驗結果顯示，採用顏色空間特徵進行歐基里德(Euclidean)相似度比對，檢索刺青影像之查全率(recall)及查準率(precision)，均優於運用城市街距(City block)相似度及坎培拉(Canberra)相似度檢索刺青影像的作法。實驗結果可提供鑑識科學參考運用。

In the ever development of technology, humanity has joint biometrics and soft biometric traits in almost every possible methods to retrieve information. Technology has given users a much easier form of searching digital data in numerous fields such as crime, suspects, and victim identification. The common issue of overflowing information in databases is being gradually more stressful to image databases; it is tough and time consuming to search individuals from a database just by fingerprint alone. It is also beneficial and effective when searching people using unique soft biometrics such as adding tattoos as an alternative search identifier.
With this being said, in this thesis, we propose a method that uses spatial color histogram which is a method commonly used for retrieving images. We use preprocessing to attain the minimal bundle square (MBS). Using gridlines division to split images, and then use its extracted vectors to search and compare their similarities with Euclidean distance. The experiment results show that the recall and precision are better than the results of using identity features with city block distance and Canberra distance. This alternative form of image retrieval can surely define a particular image and also it can facilitate forensics.

Abstract iv
Table of content vi
List of tables vii
List of figures viii
1.Introduction 1
2.Related Works 4
3.Proposed Method 9
4.Experimental Results 16
5.Conclusion and Future Works 25
References 27

[1] W. Hsu, T. S. Chua, and H. H. Pung. “An integrated color-spatial approach to content-based image retrieval”, Proceedings of the 3rd ACM international conference on multimedia, 1995, pp. 305-313.
[2] J.-E. Lee, R. Jin, A. K. Jain and W. Jong. “Image Retrieval in Forensics: Application to Tattoo Image Database”, IEEE multimedia, Vol. 19, 2012, pp. 40-49.
[3] A. K. Jain, J.-E Lee, R. Jin and N. Gregg. “Content-based image retrieval: An application to tattoo images”. Image Processing (ICIP), 16th IEEE international conference, Vol. 16, 2009, pp. 2745-2748.
[4] A. K. Jain, J.-E Lee and R. Jin, “Tattoo-ID: Automation Tattoo Image Retrieval for Suspect and Victim Identification”, In proceedings of the multimedia 8th pacific rim conference on advances in multimedia information processing, 2007, pp. 256-265.
[5] D. G. Lowe, “Distinctive image features from scale invariant keypoints”, International Journal of Computer Vision, Vol. 60, 2004, pp. 91-110.
[6] S. T. Acton and A. Rossi, “Matching and Retrieval of Tattoo Images: Active Contour CBIR and Glocal Image Features”, In proceedings of IEEE Southwest symposium on image analysis and interpretation”, 2008, pp. 21-24.
[7] A. K. Jain and A. Vailaya, “Image Retrieval using Color and Shape”, Vol. 29, 1995, pp. 1233-124.
[8] J.-E. Lee, R. Jin and A. K. Jain, “Unsupervised Ranking: Application to Large-Scale Image Retrieval”, In proceedings of (ICPR), 20th international conference, 2010, pp. 3902-3096.
[9] M. Analoui, “Content-based image Retrieval Using Artificial Immune System (AIS) Clustering Algorithm”, In proceedings of the international multimedia conference of engineers and computer scientists 2011, Vol. 1, 2011, IMGECS, India.
[10] Jain, A.K, Jung-Eun Lee, R. Jin, and N. Gregg. “Content-based image retrieval: An application to tattoo images”. Processing (ICIP), 16th 2009 IEEE International Conference, 2009, pp. 2745-2748.
[11] Jung-Eun Lee, Rong Jin, Jain, A.K, Tong W. “Image Retrieval in Forensics: Application to Tattoo Image Database”, IEEE Multimedia, Vol. 19, 2011, pp. 40-49.
[12] A, Amanatiadis, V.G. Kaburlasos, A. Gasteratos, S. E. Papadakis, “Evaluation of shape descriptors for shape-based image retrieval”, IET Journals & Magazine, Vol. 5, 2011, pp. 493-499.
[13] J. Mishra, A. Sharma, K. Chaturvedi, “An Unsupervised Cluster-based Image Retrieval Algorithm using Relevance Feedback”, IJMIT International Journal of Management Information Technology, 2011, Vol. 3.
[14] D. G. Lowe, “Distinctive Image Features from Scale-Invariant Keypoints”. International Journal of Computer Vision, Vol. 60, 2004, pp. 91-110.
[15] Lee, J-E., Jain, A. K., and Jin, R., “Scars, Marks and Tattoo (SMT): Soft biometric for suspect and victim identification”. IEEE Conference, 2008, pp. 1-8.