臺灣博碩士論文加值系統

近年來隨著網路社群的日益發達，許多人開始在社群網站中用照片，來分享自己的生活瑣事，照片中會包含日常生活的人事物，因此我們希望能從中分析這些資訊。為了得到這些照片中的資訊，過去是利用文字內容來做分析，近年來因影像辨識的進步，開始使用影像技術，挖取藏在影像中的資訊。本研究是基於邊緣偵測提出新的物件猜測(object proposal)的方法，並以飲料商品為例，將候選物件提出後，利用SIFT擷取物件特徵，再以SVM做飲料商品的分類。
本研究主要分為兩大部分：物件猜測和物件辨識。首先在網路上的蒐集各種含有商品的照片及利用不同手機所拍攝含有飲料商品的照片；接著透過邊緣偵測來找出物件的邊緣，且從邊緣中找出組成物件的封閉區域，再藉由以飲料商品所統計出的特徵，過濾候選的區域，以找出有可能為物件的區域，做為候選物件；最後，將此區域做SIFT的特徵描述並做SVM的分類預測。

Nowadays as the social media progresses, more people are starting to use photosin their social media website to share the events in their daily lines. These photos often contain people and things occurring in their lives, which can be used as the data for our information analysis. In order to collect information from these photos, in the past the used information are words, but since recently the field of image detection has made an improvement, the usage of image analysis technology becomes more widely applied on this field to mine the information from the image. In this paper, the method is based on the edge detection to be as a new way of object proposal. We use the drinks products as medium to select the assumed candidates from the drink image of database, and using SIFT to derive the characteristics of the objects. After the characteristics are decided, we use SVM to categorize the drinks from their image mark.
The essay is divided into two parts, the object proposal and the object detection. First, we collect a variety of pictures containing the commodity of the drinks from the internet and the pictures taken. Then, through the method of edge detection we get the edges of the objects. As the edges are found, we look for the closed regions in the picture, and use the filter made by the characteristics got from the statistic data to filter these areas to find the possible region locating the objects marked as candidates. Finally, we apply SIFT method on these region candidates to get the characteristic description and use SVM to make predicts of classification.

口試委員會審定書 #
誌謝 i
中文摘要 ii
ABSTRACT iii
目錄 iv
圖目錄 vi
表目錄 ix
Chapter 1 緒論 1
1.1 研究背景 1
1.2 研究動機 2
1.3 研究目的 3
1.4 論文架構 3
Chapter 2 文獻探討 5
2.1 全影像搜尋 5
2.2 物件猜測( object proposal ) 8
2.3 邊緣偵測 13
2.3.1 隨機決策森林 16
2.3.2 結構化隨機森林 18
2.3.3 Structured Forests for Fast Edge Detection 20
2.4 影像局部特徵描述 21
2.4.1 SIFT( Scale-invariant feature transform ) 23
2.5 詞袋模型 28
Chapter 3 研究方法 31
3.1 演算法流程與問題定義 31
3.2 線段過濾器 32
3.3 填色 36
3.3.1 HSV色彩空間 39
3.4 物件合併 40
3.5 SVM( 支持向量機 ) 42
Chapter 4 實驗結果 46
4.1 實驗設備環境 46
4.2 系統實驗影像 47
4.3 SVM訓練 47
4.4 系統實作與結果 49
Chapter 5 結論與未來展望 57
參考文獻 59
附錄一 62
附錄二 66
附錄三 70

[1]M.-M. Cheng, Z. Zhang, W.-Y. Lin, and P. Torr, "BING: Binarized normed gradients for objectness estimation at 300fps," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2014, pp. 3286-3293.
[2]S. Romberg, L. G. Pueyo, R. Lienhart, and R. van Zwol, "Scalable logo recognition in real-world images," pp. 1-8, 2011.
[3]H. Sahbi, L. Ballan, G. Serra, and A. Del Bimbo, "Context-dependent logo matching and recognition," IEEE Transactions on Image Processing, vol. 22, pp. 1018-1031, 2013.
[4]S. Romberg and R. Lienhart, "Bundle min-hashing for logo recognition," p. 113, 2013.
[5]J. R. Uijlings, K. E. van de Sande, T. Gevers, and A. W. Smeulders, "Selective search for object recognition," International journal of computer vision, vol. 104, pp. 154-171, 2013.
[6]C. L. Zitnick and P. Dollár, "Edge boxes: Locating object proposals from edges," in European Conference on Computer Vision, 2014, pp. 391-405.
[7]P. Dollár and C. L. Zitnick, "Structured forests for fast edge detection," in Proceedings of the IEEE International Conference on Computer Vision, 2013, pp. 1841-1848.
[8]L. G. Roberts, "Machine perception of three-dimensional soups," Massachusetts Institute of Technology, 1963.
[9]N. Kazakova, M. Margala, and N. G. Durdle, "Sobel edge detection processor for a real-time volume rendering system," in Circuits and Systems, 2004. ISCAS''04. Proceedings of the 2004 International Symposium on, 2004, pp. II-913-16 Vol. 2.
[10]J. M. Prewitt, "Object enhancement and extraction," Picture processing and Psychopictorics, vol. 10, pp. 15-19, 1970.
[11]J. Canny, "A computational approach to edge detection," IEEE Transactions on pattern analysis and machine intelligence, pp. 679-698, 1986.
[12]T. K. Ho, "Random decision forests," in Document Analysis and Recognition, 1995., Proceedings of the Third International Conference on, 1995, pp. 278-282.
[13]J. J. Lim, C. L. Zitnick, and P. Dollár, "Sketch tokens: A learned mid-level representation for contour and object detection," in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2013, pp. 3158-3165.
[14]P. Geurts, D. Ernst, and L. Wehenkel, "Extremely randomized trees," Machine learning, vol. 63, pp. 3-42, 2006.
[15]P. Kontschieder, S. R. Bulo, H. Bischof, and M. Pelillo, "Structured class-labels in random forests for semantic image labelling," in 2011 International Conference on Computer Vision, 2011, pp. 2190-2197.
[16]D. G. Lowe, "Object recognition from local scale-invariant features," in Computer vision, 1999. The proceedings of the seventh IEEE international conference on, 1999, pp. 1150-1157.
[17]H. Bay, T. Tuytelaars, and L. Van Gool, "Surf: Speeded up robust features," in European conference on computer vision, 2006, pp. 404-417.
[18]L. Juan and O. Gwun, "A comparison of sift, pca-sift and surf," International Journal of Image Processing (IJIP), vol. 3, pp. 143-152, 2009.
[19]E. Rosten and T. Drummond, "Machine learning for high-speed corner detection," in European conference on computer vision, 2006, pp. 430-443.
[20]M. Calonder, V. Lepetit, C. Strecha, and P. Fua, "Brief: Binary robust independent elementary features," in European conference on computer vision, 2010, pp. 778-792.
[21]S. Leutenegger, M. Chli, and R. Y. Siegwart, "BRISK: Binary robust invariant scalable keypoints," in 2011 International conference on computer vision, 2011, pp. 2548-2555.
[22]E. Rublee, V. Rabaud, K. Konolige, and G. Bradski, "ORB: An efficient alternative to SIFT or SURF," in 2011 International conference on computer vision, 2011, pp. 2564-2571.
[23]D. G. Lowe, "Distinctive image features from scale-invariant keypoints," International journal of computer vision, vol. 60, pp. 91-110, 2004.
[24]J. Yang, Y.-G. Jiang, A. G. Hauptmann, and C.-W. Ngo, "Evaluating bag-of-visual-words representations in scene classification," in Proceedings of the international workshop on Workshop on multimedia information retrieval, 2007, pp. 197-206.
[25]Li Fei-Fei , Rob Fergus , Antonio Torralba, "Recognizing and Learning Object Categories:Year2007",CVPR2007 http://courses.cs.washington.edu/courses/cse576/08sp/lectures/CVPR2007_tutorial-Abridged.pdf