臺灣博碩士論文加值系統

摘要
本論文提出一個區域比對的方法，由使用者圈出感興趣的區域，在壓縮影像資料庫中不解壓縮情況下而找出能代表影像的特徵，並比對感興趣區域與相似壓縮影像群是否有相似或是相同的。我們設計兩個檢索流程，第一個檢索流程是找出每一組影像的中值影像來代表整組影像，再對此中值影像做特徵擷取，並比對由感興趣的區域所擷取出來的特徵是否相似。第二個檢索流程是對此中值影像取特徵影像，再對此特徵影像做特徵擷取，並比對由感興趣的區域所擷取出來的特徵是否相似。比對模組的設計採階段性相似度比對，找出感興趣區域須分割每段的區塊數。
由實驗結果證明，我們也發現檢索中值影像之特徵影像的正確率比檢索中值影像的正確率高約4 %，所以我們在對中值影像的特徵影像時，特徵擷取使用DC值檢索正確率89.52 %。每一區段的區塊數在3~5塊之間的有較佳的檢索正確率。

In this paper, a novel scheme has been proposed for image retrieval task using the features in a region of interest (ROI), extracted directly from a compressed or uncompressed image. The texture information is first extracted by exploiting the multiresolution nature of wavelet decomposition, which represents the horizontal, vertical and diagonal frequency distribution of an image. We then calculate the mean and standard deviation of wavelet coefficients of each sub-band as texture features. The retrieval strategy finds out the centroid image of the whole set of similar medical images, and then finds out the feature image aiming at the centroid images of different formats (compressed and uncompressed) to represent the set of medical images. Moreover, the features obtained from the centroid image and the feature images are used to evaluate the retrieval efficiency in the term of Recall and Precision values. According to the experimental results, we found the Precision value using the DC feature obtained from the feature image is 89.52 %, which is superior approximately 4 % than the used features from the centroid image. The experimental results show that the retrieval efficiency is considerably improved by the proposed approach.

第一章 緒論.................................1
1.1 研究背景...................................1
1.2 研究動機及目的.............................2
1.3 論文架構...................................3
第二章 研究原理.............................4
2.1 影像壓縮模型.........................4
2.1.1 JPEG.........................5
2.1.2 JPEG2000.....................7
2.2 影像檢索模型........................13
2.2.1 影像特徵擷取................14
2.2.2 影像特徵比對................14
2.3 感興趣的區域(Region of Interest)....15
2.4 相似影像群的分類方法................15
2.5 中值法..............................16
第三章 影像檢索策略...........................18
3.1 檢索流程設計..............................19
3.2 特徵擷取模組..............................21
3.2.1 影像特徵擷取............................21
3.2.2 特徵影像................................22
3.3 特徵比對方法..............................28
3.3.1 傳統比對...............................30
3.3.2 循序分段比對...........................31
3.3 效能評估方法..............................33
第四章 實驗結果與效能比較.....................34
4.1 測試影像組................................34
4.2 實做環境..................................35
4.3 檢索結果..................................36
第五章 討論與未來展望.........................58
參考文獻......................................59

[1] G. R. Shyu, C. Pavlopoulou, A. C. Kak and C. E. Brodley, “ Using Human Perceptual Categories for Content-Based Retrieval from a Medical Image Database, ” Computer Vision and Image Understanding, vol. 88, pp. 119-15, 2002.
[2] Avi Kak and Christina Pavlopoulou, “ Content-Based Retrieval from a Medical Image Database, ” Pattern Recognition, pp. 119-152002.
[3] G. Qiu, “Indexing chromatic and achromatic patterns for content-based colour image retrieval, ” Pattern Recognition, vol. 35, pp. 1675-1686, 2002.
[4] N. Nikolaoua and N. Papamarkosb, “Color image retrieval using a fractal signature extraction technique, ” Engineering Applications of Artificial Intelligence, vol. 15, pp. 81-96, 2002.
[5] B. S. Manjunath and W. Y. Ma, “ Texture Features for Browsing and Retrieval of Image Data, ”IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 18, pp. 837-842 , 1996.
[6] Y. K. Chan, C. C. Chang, “A fast filter of spatial video retrieved,”Applied Mathematics, pp. 157—171, 2000.
[7] J. Jiang, A. Armstrong, G. Feng, “ Direct content access and extraction
from JPEG compressed images, ” Pattern Recognition, val. 35, pp. 2511—2519, 2002.
[8] G. Schaefer, “JPEG image retrieval by simple operators, ” Proceedings of Second International Workshop on Content Based Multimedia and Indexing, pp. 207—214, 2001.
[9] G. Feng, “JPEG compressed image retrieval via statistical features, ” Pattern Recognition, vol. 36, pp. 977-985, 2003.
[10] J. Puniene, V. Punys and J. Punys, “Ultrasound and angio image comprssion by cosine and wavelet transforms, ” Medical Informatics, vol. 64, pp. 473-481, 2001.
[11] J. H. Liu and H. Gu, “Image retrieval in various domains, ” Computers & Graphics, vol. 27, pp. 807-812, 2003.
[12]J. Z. Wang, G. Wiederhold, O. Firschein and S. X. Wei, “Content-based image and Daubechies’ wavelets, ”Digital Libraries, pp. 331-328, 1997.*/
[13] S. Arivazhagan, L. Ganesan, “Texture classification using wavelet transform, ” Pattern Recignition, vol. 24, pp. 1513-1521, 2003.
[14] N. Suematsu, Y. Ishida, A. Hayashi and T. Kanbara, “Region-based Image Retrieval using Wevelet Transform,” Image and Vision Computing, 2002.
[15] Sharlee Climer and Sanjiv K. Bhatia, “ Image database indexing using JPEG coefficients, ” Pattern Recognition, vol. 35 , pp.2479-2488, 2002.
[16] A. Agarwal, A. H. Rowberg, and Y. Kim, “Fast JPEG 2000 Decoder and Its Use in Medical Imaging, ” IEEE Transactions on Information Technology in Biomedicine, vol. 7, pp. 184-190, 2003.
[17] X. L. Li, “Image retrieval based on perceptive weighted color blocks, ” Pattern Recognition, vol. 24, pp. 1935-1941, 2003.
[18] Bo Tao and Bradley W. Dickinson, “ Texture Recognition and Image Retrieval Using Gradient Indexing, ” Visual Communication and Image Representation, vol. 11, pp. 327-342, 2000.
[19] Y. K. Chan, C. C. Chang, “Image matching using run-length feature, ”Pattern Recognition, vol. 22, pp. 447—455, 2001.
[20] R. Brunelli, O. Mich, C. M. Modena, “A survey on the automatic indexing of video data,” Visual Communicationand Image Representation, vol. 10 , pp. 78—112, 1999.
[21] Y. K. Chan, C. C. Chang, “A fast filter of spatial video retrieved,”Applied Mathematics, pp. 157—171, 2000.