臺灣博碩士論文加值系統

本論文提出一套具有性別辨識功能的無人自走車系統，此系統可以在公眾環境中對行人進行偵測與追蹤，並進一步進行性別辨識，以期能夠針對不同性別提供相關資訊。
本系統架構可分為兩個部分，分別為自走車的行人偵測追蹤，以及針對目標進行性別辨識。首先在目標偵測及追蹤部分，本系統利用AdaBoost演算法進行目標偵測，再藉由KLT追蹤演算法追蹤目標位置，並以Microsoft Kinect Sensor作為偵測目標距離以及判斷行進是否停止的輔助裝置。在性別辨識部分，我們結合多尺度區域二元特徵（Local Binary Patterns, LBP）及方向導數直方圖（Histograms of Oriented Gradients, HOG）做為性別分類的主要特徵，並搭載PSO-SVM(Particle Swam Optimization-Support Vector Machines)分類演算法，我們將特徵以統計學上用於計算平均差的t-test方式，以計算所得的p-value來進行特徵篩選，使得我們的特徵可以更有效益的結合，在準確率仍達92.6%的情況下，但特徵數量可以大幅減少至原有特徵量的30%，使運算速度提升。

In this study, an automatic guidance vehicle (AGV) service robot system with gender recognition function is proposed. This system can detect and track a pedestrian in front of AGV. Moreover, the gender of the target can be recognized, thus the AGV can provide different information according to the target’s gender.
The proposed system can be divided into two parts: the first part is the pedestrian detection and tracking function by the AGV robot, the second part is the gender recognition. In the part of pedestrian detection and tracking, AdaBoost algorithm is used to detect the target, and the Kanade-Locus-Tomasi feature tracker is then applied to track the position of the target. Microsoft Kinect Sensor is also utilized as an auxiliary device to measure distance to the target and to determine whether the AGV should stop or not. In the gender recognition part, the proposed system combines multi-scale Local Binary Patterns (LBP) and Histograms of Oriented Gradients (HOG) for feature extraction, and the Particle Swam Optimization with Support Vector Machines (PSO-SVM) is chosen as the classification algorithm. After the feature extraction, the t-test is used to calculate p-value for each feature to determine if it has significant difference between male and female categories. Next, we train SVM model with the significant features, which are selected by the p-value. Experimental results show that the classification accuracy is up to 92.6%, while only 30% of the original features are used.

目錄
中文摘要 i
英文摘要 ii
圖目錄 v
表目錄 viii
第一章緒論 1
1.1研究背景與動機 1
1.2研究目的 3
1.3文獻回顧 4
1.4論文架構 7
第二章系統架構 9
2.1物件偵測 10
2.2 Kanade-Lucas-Tomasi追蹤演算法 14
2.3 Microsoft Kinect感測裝置 16
第三章性別辨識 18
3.1區域二元特徵(Local Binary Patterns) 19
3.2方向導數直方圖(Histograms of Oriented Gradients) 22
3.3區域二元特徵及方向導數直方圖的多尺度（multi-scale）特徵擷取 24
3.4 T-test 26
3.5支援向量機(Support Vector Machines) 33
3.6粒子群聚演算法(Particle Swam Optimization) 36
第四章實驗結果與討論 40
4.1硬體規格 40
4.2人臉資料庫與前處理 44
4.3以P-value篩選多尺度區域二元特徵及方向導數直方圖的結果 49
4.4系統實際測試 55
第五章結論與未來展望 62


圖目錄
圖 1-1 打掃型機器人以及機器狗 2
圖 1-2 好奇號火星探測車 2
圖 1-3 PCA產生的Eigenface 5
圖 2-1 機器人主要行為流程圖 10
圖 2-2 海爾特徵 11
圖 2-3 點處的積分影像面積 13
圖 2-4 積分影像運算 14
圖 2-5 追蹤影像 15
圖 2-6 Kinect深度資訊影像 16
圖 3-1 性別辨識流程圖 18
圖 3-2 LBP 20
圖 3-3 uniform pattern 21
圖 3-4 LBP切割意識圖 21
圖 3-5 HOG執行後結果。由左至右依序為1.原始影像2.X方向微分影像3.Y方向微分影像4.純量m影像 23
圖 3-6 半徑為1的尺度 24
圖 3-7 半徑為2的尺度 25
圖 3-8 半徑為2的尺度運算 25
圖 3-9 不同群組的平均值差異情形 27
圖 3-10 具有相同平均值差異不同變異數的兩種分佈 28
圖 3-11 不同平均值差異C的T分佈情形 29
圖 3-12 P值表示圖 31
圖 3-13 SVM切割圖 33
圖 3-14 線性不可切割示意圖 34
圖 3-15 PSO演算法流程圖 37
圖 4-1 機器人硬體架構 40
圖 4-2 PTZ攝影機 41
圖 4-3 Pioneer 3-DX 42
圖 4-4 Kinect 43
圖 4-5 Aberdeen示意圖 45
圖 4-6 Frontal face dataset示意圖 45
圖 4-7 GUFD示意圖 46
圖 4-8 Pain expressions示意圖 46
圖 4-9 Utrecht ECVP示意圖 47
圖 4-10 人臉偵測圖 47
圖 4-11人臉五官偵測圖 48
圖 4-12 切割影像 48
圖 4-13直方圖等化結果 49
圖 4-14 P-value對應準確率 52
圖 4-15 P-value對應特徵比例 53
圖 4-16 攝影機辨識結果 56
圖 4-17 系統行人偵測 56
圖 4-18以行人為目標進行追蹤 57
圖 4-19追蹤目標位移 58
圖 4-20 將位移目標修正回攝影機中間 59
圖 4-21行進間對目標進行辨識 60
圖 4-22目標前顯示辨識結果 61





表目錄
表4- 1 Canon PTZ規格明細表 41
表4- 2 Pioneer 3-DX規格明細表 42
表4-3 函式庫整理說明 44
表4-4 Aberdeen資料庫明細 45
表4-5 Frontal face dataset資料庫明細 45
表4-6 GUFD資料庫明細 46
表4-7 Pain expressions資料庫明細 46
表4-8 Utrecht ECVP資料庫明細 47
表4-9多尺度的LBP 50
表4-10多尺度的HOG 50
表4-11特徵經P-value篩選結果 51
表4-12特徵經P-value篩選對時間的影響 54
表4-13 LBP(R=4)結合HOG(R=3)篩選後的特徵數量與準確率 54
表4-14 LBP(R=1~8)+HOG(R=1~8)篩選後的特徵數量與準確率 55

[1]http://www.roomba.com.tw/product/780.html
[2]http://www.sony.net/SonyInfo/News/Press_Archive/199905/99-046/index.html
[3]http://www.nasa.gov/mission_pages/msl/index.html
[4]B. Yamauchi, A. Shultz and W. Adams, “Mobile robot exploration and map-building with continuous localization,” in Proceedings of the Internet Content Rating Association (ICRA), vol. 4, pp. 3175-3720,1998.
[5]P. Vadakkepat, P. Lim, L. Silva, L. Jing, L. Ling ”Multimodal approach to human-face detection and tracking,” IEEE Transactions on Industrial Electronics,vol. 55, no. 3, pp.1385-1393, 2008
[6] B. Moghaddam and M.H. Yang, “Learning Gender with Support Faces,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 24, no. 5, pp. 707-711, 2002.
[7] K. Balci and V. Atalay, “PCA for Gender Estimation: Which Eigenvectors ontribute?,” in Proceedings of The 16th International Conference on Pattern Recognition, vol. 3, pp. 363-366, 2002.
[8]李秉翰, “電腦視覺監控產學研聯盟電子報,”電腦視覺監控產學研聯盟,第七期, Jan. 2006.
[9] H.C. Lian and B.L. Lu, “Multi-view Gender Classification Using Multi-Resolution Local Binary Patterns and Support Vector Machines,” International Journal of Neural Systems, vol. 17, no. 6, pp. 479-487, 2007.
[10] J.S.O. Dniz, G. Bueno, and F.D. la Torre, “Face Recognition Using Histograms of Oriented Gradients,” Pattern Recognition Letters, vol. 32, no. 12, pp. 1598-1603, 2011.
[11] B. Xia, H. Sun, and B.L. Lu, “Multi-view Gender Classification based on Local Gabor binary Mapping Pattern and Support Vector Machines,” in Proceedings of IEEE International Joint Conference on Neural Networks, pp. 3388-3395, 2008.
[12] C. Shan, “Learning Local Binary Patterns for Gender Classification on Real-World Face Images,” Pattern Recognition Letter, vol. 33, no. 4, pp. 431-437, 2011.
[13] P. Viola and M.J. Jones, “Rapid Object Detection using a Boosted Cascade of Simple Features,”in Proceedings of the IEEE Computer Society International Conference on Computer Vision and Pattern Recognition, vol. 1, pp. 511-518, Dec., 2001.
[14] C. Tomasi and T. Kanade, “Detection and tracking of point features,” International Journal of Computer Vision, 1991
[15] T. Ojala, M. Pietikainen and T. Maenpaa,“ Multiresolution Gray-Scale and Rotation Invariant Texture Classification with Local Binary Patterns,”IEEE Trans. Pattern Analysis and Machine Intelligence,vol. 24, no. 7, pp. 971-987, Jul., 2002.
[16] W. Ronald, M. Raymond, M. Sharon and Y. Keying, “Probability and Statistics for Engineers and Scientists,” Pearson, pp. 237, 2002.
[17] J. Kennedy and R. C. Eberhart, “Particle swarm optimization,” in Proceedings of IEEE International Conference on Neural Networks, vol. 4, pp. 1942-1948, 1995
[18]OpenCV, http://www.opencv.org/
[19]OpenNI, http://openni.org/
[20]Aria,http://robots.mobilerobots.com/
[21] http://pics.psych.stir.ac.uk/zips/Aberdeen.zip
[22] http://www.robots.ox.ac.uk/~vgg/data/data-cats.html
[23]http://homepages.abdn.ac.uk/m.burton/pages/gfmt/Glasgow%20Face%20Recognition%20Group.html
[24] http://pics.psych.stir.ac.uk/zips/pain.zip
[25] http://pics.psych.stir.ac.uk/zips/utrecht.zip