基於殘差神經網路之比特犬辨識

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：50

、訪客IP：18.116.49.38

姓名

鍾裕廷(YU-TING CHUNG) 查詢紙本館藏

畢業系所

資訊工程學系在職專班

論文名稱

基於殘差神經網路之比特犬辨識
(Residual Neural Network Based Recognition in Pit Bull-Type Dogs)

相關論文

★ 整合GRAFCET虛擬機器的智慧型控制器開發平台	★ 分散式工業電子看板網路系統設計與實作
★ 設計與實作一個基於雙攝影機視覺系統的雙點觸控螢幕	★ 智慧型機器人的嵌入式計算平台
★ 一個即時移動物偵測與追蹤的嵌入式系統	★ 一個固態硬碟的多處理器架構與分散式控制演算法
★ 基於立體視覺手勢辨識的人機互動系統	★ 整合仿生智慧行為控制的機器人系統晶片設計
★ 嵌入式無線影像感測網路的設計與實作	★ 以雙核心處理器為基礎之車牌辨識系統
★ 基於立體視覺的連續三維手勢辨識	★ 微型、超低功耗無線感測網路控制器設計與硬體實作
★ 串流影像之即時人臉偵測、追蹤與辨識─嵌入式系統設計	★ 一個快速立體視覺系統的嵌入式硬體設計
★ 即時連續影像接合系統設計與實作	★ 基於雙核心平台的嵌入式步態辨識系統

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 (2026-10-1以後開放)

摘要(中)

由於比特犬的攻擊事件頻傳，比特犬其特有的生理條件以及長期被培育成殘忍的鬥犬，因此傷人或傷犬的事件時有耳聞，受害者等特定團體因而要求政府全面禁養，目前世界上有數十個國家已將比特犬列為管制犬種，而台灣也將與國際接軌，目前已預告將比特犬列為管制輸入，禁止從其他國家進口到台灣，已飼養的比特犬雖可繼續飼養，但需要登記與列冊管理，而不得任意繁殖與販售。
要從源頭管理禁止輸入比特犬，則必須建立完善邊境管理方式，目前一般國際間的海關通常是以外觀作為犬種鑑定，但由於人的肉眼判斷標準不一致，容易出現爭議，若能以犬隻影像建置一套辨識系統，則能夠避免人為判斷的爭議，故本研究希望能以AI方式辨識比特犬，以動物的生理特徵結合電腦視覺來辨識，在訓練過程中需要大量的比特犬影像，首先將影像去除背景擷取出比特犬的影樣，接著以三種神經網路辨識比特犬，分別是殘差神經網路、孿生神經網路以及殘差神經網路與孿生神經網路組合的方式，三種實驗結果顯示以殘差神經網路的方式辨識比特犬的效果最佳。

摘要(英)

Due to frequent attacks of pit bulls, pit bulls have special physical conditions and long-term bred to be cruel fighting dogs. Incidents of hurting people or dogs are heard sometimes, some groups of victims require the government to ban pit bulls. At present, many countries in the world have listed pit bulls as a controlled dog breed, Taiwan will be in line with international standards. It has been announced that Taiwan bans pit bulls imports from other countries. Those pit bulls have been bred that can be bred, but they need to be registered and managed. Pit bulls cannot be arbitrarily bred and sold in the future.
To efficiently ban the import of pit bulls from other countries, it is necessary to establish border management methods. At present, international customs usually use appearance as the recognition of dog breeds, human judgment is inconsistent and may cause controversies. The controversies of human judgment can be avoided if we establish efficient recognition system with pit pull-type dogs′ image. Therefore, this study will use AI biometric identification to recognize pit bulls with computer vision, a large number of pit bull images are needed in the training process. The background image is removed and the dog image is cropped, and use three different neural networks to recognize pit bulls. There are Deep Residual Network, Siamese Network and the above network combination method. The best result of three methods is to use the Deep Residual Network to recognize pit bull.

關鍵字(中)

★ 深度學習
★ 殘差神經網路
★ 比特犬

關鍵字(英)

論文目次

中文摘要 i
Abstract ii
誌謝辭 iii
目錄 iv
圖目錄 vii
表目錄 ix
第1章緒論 1
1-1 研究背景 1
1-2 研究目的 2
1-3 論文架構 3
第2章文獻回顧 5
2-1 卷積神經網路 5
2-2 物件偵測與YOLO簡介 8
2-3 殘差神經網路簡介 14
2-4 孿生神經網路簡介 16
第3章比特犬辨識方法與系統設計 18
3-1 MIAT高階系統設計方法論 18
3-1-1 IDEF0階層架構 19
3-1-2 Grafcet離散事件建模 20
3-2 比特犬辨識方法與系統 21
3-3 YOLO犬隻偵測與切割模組設計 22
3-4 殘差神經網路分類模組設計 24
3-5 孿生神經網路決策模組設計 26
第4章實驗結果與分析 28
4-1 實驗平台 28
4-2 實驗資料集 29
4-3 YOLO物件偵測與切割實驗 32
4-4 殘差神經網路辨識實驗 33
4-4-1 比特犬類與非比特犬類ResNet實驗 33
4-4-2 管制犬類與非管制犬類ResNet實驗 36
4-5 孿生神經網路辨識實驗 39
4-5-1 比特犬類與非比特犬類Siamese實驗 39
4-5-2 管制犬類與非管制犬類Siamese實驗 41
4-6 實驗結果比較 42
4-7 實驗呈現 43
4-7-1 比特犬類與非比特犬類實驗結果 43
4-7-2 管制犬類與非管制犬類實驗結果 45
4-8 結論 47
參考文獻 48

參考文獻

[1] K. R. Olson, J. K. Levy, B. Norby, M. M. Crandall, J. E. Broadhurst, S. Jacks, R. C. Barton, and M. S. Zimmerman, "Inconsistent identification of pit bull-type dogs by shelter staff," The Veterinary Journal, vol. 206, no. 2, pp. 197-202, 2015.
[2] Y. LeCun, Y. Bengio, and G. Hinton, "Deep learning," nature, vol. 521, no. 7553, pp. 436-444, 2015.
[3] C. E. Bugge, J. Burkhardt, K. S. Dugstad, T. B. Enger, M. Kasprzycka, A. Kleinauskas, M. Myhre, K. Scheffler, S. Ström, and S. Vetlesen, "Biometric methods of animal identification," Course notes, Laboratory Animal Science at the Norwegian School of Veterinary Science, pp. 1-6, 2011.
[4] A. Anuntachai and N. Pantuwong, "An Image-based Sea Turtle Identification using Postorbital Facial Feature Points Matching Technique," in 2019 19th International Conference on Control, Automation and Systems (ICCAS), pp. 1058-1063, 2019.
[5] R. Maglietta, R. Caccioppoli, E. Seller, S. Bellomo, F. C. Santacesaria, R. Colella, G. Cipriano, E. Stella, K, Hartman, and C. Fanizza, "Convolutional neural networks for Risso′s dolphins identification," IEEE Access, vol. 8, pp. 80195-80206, 2020.
[6] E. Ranguelova, M. Huiskes, and E. J. Pauwels, "Towards computer-assisted photo-identification of humpback whales," in 2004 International Conference on Image Processing (ICIP′04), vol. 3, pp. 1727-1730, 2004.
[7] S. Kumar and S. K. Singh, "Biometric recognition for pet animal," Journal of Software Engineering and Applications, vol. 7, no. 5, pp. 470-482, 2014.
[8] M. Chanvichitkul, P. Kumhom, and K. Chamnongthai, "Face recognition based dog breed classification using coarse-to-fine concept and PCA," in 2007 Asia-Pacific Conference on Communications, pp. 25-29, 2007.
[9] G. Mougeot, D. Li, and S. Jia, "A Deep Learning Approach for Dog Face Verification and Recognition," in Pacific Rim International Conference on Artificial Intelligence, pp. 418-430, 2019.
[10] W. Lu, "Dog′s nose segmentation and nose-print pet identity recognition," National Central University, 2020.
[11] Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner, "Gradient-based learning applied to document recognition," Proceedings of the IEEE, vol. 86, no. 11, pp. 2278-2324, 1998.
[12] R. Girshick, J. Donahue, T. Darrell, and J. Malik, "Rich feature hierarchies for accurate object detection and semantic segmentation," in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 580-587, 2014.
[13] R. Girshick, "Fast r-cnn," in Proceedings of the IEEE international conference on computer vision, pp. 1440-1448, 2015.
[14] S. Ren, K. He, R. Girshick, and J. Sun, "Faster r-cnn: Towards real-time object detection with region proposal networks," arXiv preprint arXiv:1506.01497, 2015.
[15] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You only look once: Unified, real-time object detection," in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 779-788, 2016.
[16] W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C. Y. Fu, and A. C. Berg, "Ssd: Single shot multibox detector," in European conference on computer vision (ECCV 2016), pp. 21-37, 2016.
[17] J. Redmon and A. Farhadi, "YOLO9000: better, faster, stronger," in Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR), pp. 7263-7271, 2017.
[18] J. Redmon and A. Farhadi, "Yolov3: An incremental improvement," arXiv preprint arXiv:1804.02767, 2018.
[19] A. Bochkovskiy, C. Wang, and H. Liao, "Yolov4: Optimal speed and accuracy of object detection," arXiv preprint arXiv:2004.10934, 2020.
[20] D. Misra, "Mish: A self regularized non-monotonic neural activation function," arXiv preprint arXiv:1908.08681, 2019.
[21] Y. Bengio, P. Simard, and P. Frasconi, "Learning long-term dependencies with gradient descent is difficult," IEEE Transactions on Neural Networks, vol. 5, no. 2, pp. 157-166, 1994.
[22] X. Glorot and Y. Bengio, "Understanding the difficulty of trainingdeep feedforward neural networks," in Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, pp. 249-256, 2010.
[23] I. J. Goodfellow, D. Warde-Farley, M. Mirza, A. Courville, and Y. Bengio, "Maxout networks," arXiv preprint arXiv:1302.4389, 2013.
[24] G. E. Hinton, N. Srivastava, A. Krizhevsky, I. Sutskever, and R. R. Salakhutdinov, "Improving neural networks by preventing co-adaptation of feature detectors," arXiv preprint arXiv:1207.0580, 2012.
[25] S. Ioffe and C. Szegedy, "Batch normalization: Accelerating deep network training by reducing internal covariate shift," arXiv preprint arXiv:1502.03167, 2015.
[26] K. He, X. Zhang, S. Ren, and J. Sun, "Deep residual learning for image recognition," in Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR), pp. 770-778, 2016.
[27] J. Bromley, I. Guyon, Y. LeCun, E. Sckinger and R. Shah, "Signature Verification using a "Siamese" Time Delay Neural Network," Proceedings of the 7th Annual Neural Information Processing Systems, vol. 7, no. 4, pp. 669-688, 1993.
[28] C. Chen, C. Kuo, C. Chen, and J. Dai, "The design and synthesis using hierarchical robotic discrete-event modeling," Journal of Vibration and Control, vol. 19, pp. 1603-1613, 2013.
[29] R. David, "Grafcet: a powerful tool for specification of logic controllers," IEEE Transactions on Control Systems Technology, vol. 3, no. 3, pp. 253-268, 1995.
[30] S. Ioffe and C. Szegedy, "Batch normalization: Accelerating deep network training by reducing internal covariate shift," arXiv preprint arXiv:1502.03167, 2015.

指導教授

陳慶瀚(CHING-HAN CHEN)

審核日期

2021-10-29

推文