請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/17620
完整後設資料紀錄
DC 欄位 | 值 | 語言 |
---|---|---|
dc.contributor.advisor | 楊宏智(Hong-Tsu Young) | |
dc.contributor.author | Jauh-Hsiang Lan | en |
dc.contributor.author | 藍兆祥 | zh_TW |
dc.date.accessioned | 2021-06-08T00:25:49Z | - |
dc.date.copyright | 2021-02-22 | |
dc.date.issued | 2021 | |
dc.date.submitted | 2021-02-04 | |
dc.identifier.citation | [1] 科技報橘, '智慧工廠背後的管理深思:重點不是自動化,而是工廠的「統一大腦」,' July 2018. [Online] https://buzzorange.com/techorange/2018/07/30/ factory-in-the-future/. [Accessed 12 01 2020] [2] Jung Min-hee, 'Smart Factories Improving Productivity of SMEs,' March 2016. [Online] http://www.businesskorea.co.kr/news/articleView.html?idxno=14073 . [Accessed 12 01 2020] [3] Philip Liu, 'Smart Manufacturing Gains Ground in Taiwan,' December 2019. [Online] https://topics.amcham.com.tw/2019/12/smart-manufacturing-taiwan/ [Accessed 12 01 2020] [4] Nvidia, 'GPU 運算領導者,' [Online] https://www.nvidia.com/zh-tw/about-nvidia/ai-computing/ [Acessed 12 01 2020] [5] W. S. McCulloch and W. Pitts, 'A logical calculus of the ideas immanent in nervous activity,' The bulletin of mathematical biophysics, Vol.5, 1943. [6] Y. Lecun, L. Bottou, Y. Bengio and P. Haffner, 'Gradient-based learning applied to document recognition,' in Proceedings of the IEEE, vol. 86, no. 11, pp. 2278-2324, November 1998. [7] A. Krizhevsky, I. Sutskever and G. E. Hinton, 'Imagenet classification with deep convolutional neural networks,' Advances in neural information processing systems, 2012. [8] 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, 2013. [9] R. Girshick, 'Fast R-CNN,' in Proceedings of the IEEE international conference on computer vision, pp. 1440-1448, 2015. [10] S. Ren, K. He, R. Girshick and J. Sun, 'Faster R-CNN: Towards real-time object detection with region proposal networks,' IEEE transactions on pattern analysis and machine intelligence, 39(6), 1137-1149, 2016. [11] The Business Research Company, 'Wood Processing Market Global Briefing 2018,' January 2018. [Online] https://www.thebusinessresearchcompany.com/report/wood-processing-market-global-briefing-2018 [Acessed 12 01 2020] [12] Grand View Research, 'Wood Based Panel Market Size, Share Trends Analysis Report By Product (Plywood, MDF, HDF, OSB, Particleboard, Hardboard), By Application (Furniture, Construction), By Region, And Segment Forecasts, 2020 – 2027,' March 2020. [Online] https://www.grandviewresearch.com/industry-analysis/wood-based-panel-market [Acessed 12 01 2020] [13] Sveza, 'Catalog of Plywood Grades,' September 2017. [Online] https://www.sveza.com/upload/ib lock/ab f/ab faebad 008 e050ab476ed8ad abd4 c9c.p df [Acessed 12 01 2020] [14] Y. Gu, H. Andersson and R. Vicen, 'Automatic classification of wood defects using support vector machines, ' In International Conference on Computer Vision and Graphics, pp. 356-367, November 2008. [15] Hinton, G. E., Salakhutdinov, R. R., 'Reducing the dimensionality of data with neural networks. ' science, 313(5786), 504-507, 2006 [16] Sagar Sharma, 'Activation Functions in Neural Networks,' September 2017. [Online] https://towardsdatascience.com/activation-functions-neural-networks- 1cbd9f8d91d6. [Accessed 01 15 2021] [17] Shivani Kolungade, 'Object Detection, Image Classification and Semantic Segmentation using AWS Sagemaker,' August 2020. [Online] https://medium. com/@kolungade.s/object -detection -image-classification -and-semantic-segmentation-using-aws-sagemaker-e1f768c8f57d [Accessed 01 15 2021] [18] Simonyan, K., Zisserman, A., 'Very deep convolutional networks for large-scale image recognition, ' arXiv preprint arXiv:1409.1556, 2014 [19] He, K., Zhang, X., Ren, S., Sun, J., 'Deep residual learning for image recognition. ' In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770-778, 2016 [20] Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., ... Rabinovich, A., 'Going deeper with convolutions. ' In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1-9, 2015. [21] Ioffe, S., Szegedy, C., 'Batch normalization: Accelerating deep network training by reducing internal covariate shift. ' In International conference on machine learning, pp. 448-456, June 2015. [22] Long, J., Shelhamer, E., Darrell, T., 'Fully convolutional networks for semantic segmentation,' In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 3431-3440, 2015. [23] Ronneberger, O., Fischer, P., Brox, T., 'U-net: Convolutional networks for biomedical image segmentation. ' In International Conference on Medical image computing and computer-assisted intervention, pp. 234-241, October 2015. [24] He, K., Gkioxari, G., Dollár, P., Girshick, R., 'Mask r-cnn. ' In Proceedings of the IEEE international conference on computer vision, pp. 2961-2969, 2017. [25] Redmon, J., Divvala, S., Girshick, R., Farhadi, A., '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. [26] Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C. Y., Berg, A. C., 'Ssd: Single shot multibox detector. ' In European conference on computer vision, pp. 21-37, October 2016. [27] Huang, J., Rathod, V., Sun, C., Zhu, M., Korattikara, A., Fathi, A., ... Murphy, K., 'Speed/accuracy trade-offs for modern convolutional object detectors. ' In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 7310-7311, 2017. [28] Kingma, D. P., Welling, M., 'Auto-encoding variational bayes.' arXiv preprint arXiv:1312.6114, 2013. [29] Juliendespois, 'Autoencoders - Deep Learning bits #1,' February 2017. [Online] https://hackernoon.com/autoencoders -deep-learning-bits-1-11731e200694. [Accessed 01 15 2021] [30] Goodfellow, I. J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., ... Bengio, Y., 'Generative adversarial networks, ' arXiv preprint arXiv:1406.2661, 2014. [31] Mirza, M., Osindero, S., 'Conditional generative adversarial nets.' arXiv preprint arXiv:1411.1784, 2014. [32] Zoph, B., Cubuk, E. D., Ghiasi, G., Lin, T. Y., Shlens, J., Le, Q. V., 'Learning data augmentation strategies for object detection. ' In European Conference on Computer Vision, pp. 566-583, August 2020. [33] Isola, P., Zhu, J. Y., Zhou, T., Efros, A. A., 'Image-to-image translation with conditional adversarial networks. ' In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1125-1134, 2017. [34] Park, T., Liu, M. Y., Wang, T. C., Zhu, J. Y., 'Semantic image synthesis with spatially-adaptive normalization. ' In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 2337-2346, 2019. [35] Teledyne DALSA, 'Camera Models.' [Online] https://www.teledynedalsa.com/ en/products/imaging/cameras/linea/ [Acessed 12 25 2020] [36] Chiopt, '低倍率線掃描鏡頭.' [Online] http://www.chiopt.com/product/17.html [Acessed 12 25 2020] [37] Bestintech, ' HILL 高亮度線光源.' [Online] http://www.bestintech.com.tw/zh-tw/products-detail/id/41/title/HILL_High_Power_Linear_Lights [Acessed 12 25 2020] [38] Tzutalin, 'LabelImg,' Git code, 2015. [Online] https://github.com/tzutalin/labelImg [Acessed 12 25 2020] [39] Nvidia, 'GEFORCE® GTX 1080 Ti.' [Online] https://www.nvidia.com/zh-tw/geforce/products/10series/geforce-gtx-1080-ti/ [Acessed 12 25 2020] [ 4 0 ] Te n s o r f l ow, ' Te n s o r Bo a r d . ' [ On l i n e ] h t t p s : / / www. t e n s o r f l ow. o r g / tensorboard?hl=zh-tw [Acessed 12 25 2020] [41] Tensorflow, ' object_detection,' Git code, 2017. [Online] https://github.com /tensorflow/models/tree/master/research/object_detection [Acessed 12 25 2020] [42] Phillipi, 'pix2pix,' Git code, 2017. [Online] https://github.com/phillipi/pix2pix [Acessed 12 25 2020] [43] Taki0112, 'SPADE-Tensorflow,' Git code, 2017. [Online] https://github.com/ taki0112/SPADE-Tensorflow [Acessed 12 25 2020] [44] Everingham, M., Eslami, S. A., Van Gool, L., Williams, C. K., Winn, J., Zisserman, A., ' The pascal visual object classes challenge: A retrospective, ' In International journal of computer vision, 111(1), pp. 98-136, 2015. [45] Lin, T. Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., ... Zitnick, C. L., ' Microsoft coco: Common objects in context, ' In European conference on computer vision, pp. 740-755, September 2014. | |
dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/17620 | - |
dc.description.abstract | 隨著近年工業4.0的風潮漸起,如何將智慧製造應用於產線中成為產業界十分重視的課題。現今在木板、木材貼皮等木製產品加工產線中,產品瑕疵檢測大部分仍由人工進行,產品瑕疵檢測作為產線中重要的一環,如能導入自動化光學檢測將是邁向產線智慧化很好的一個著力點。然而,對於木材表面這類外觀組成複雜且無規律的檢測目標,較難以傳統電腦視覺演算法找出其中瑕疵。伴隨人工智慧深度學習技術的飛速發展,現今已有多種物件偵測模型被提出,此類模型具有高適應性與可在雜亂背景中找出目標的能力,以其作為全新的木材表面瑕疵檢測手段極具潛力。然而,物件偵測模型需使用大量影像訓練才有較好的辨識效果,在實務經驗中,常常遇到產品樣本數過少導致檢測效果不理想的情況,如能克服此問題,將大幅提升將物件偵測模型用於產線的可能性。 本研究將實際應用物件偵測模型檢測木材瑕疵,評估其是否適合作為木材表面瑕疵檢測的新方法,並探討使用不同特徵擷取器對於模型檢測效果的差異。此外本研究將使用影像生成模型生成木材瑕疵影像,測試在原始樣本數較少的情況下加入生成影像一起訓練是否能提升物件偵測模型的瑕疵檢測效果。最後根據結果提出一套針對少樣本情況下的木材瑕疵檢測流程。 經過本研究的實驗測試,物件偵測模型Faster R-CNN可用於木材瑕疵檢測,且在樣本數量較為充足的情況下,不論使用VGG16、Resnet50、InceptionV2作為特徵擷取器,其都有近六成的瑕疵檢出率。在少樣本的情況下,加入影像生成模型Pix2Pix或SPADE生成的瑕疵影像一起訓練後將提升模型瑕疵檢測能力,其中,使用InceptionV2作為特徵擷取器加上SPADE生成的瑕疵影像進行訓練時,其能將瑕疵檢出率從原本的41.58%提升至89.34%,改善效果十分顯著,以此組合作為少樣本情況下木材表面瑕疵的檢測手段將非常有潛力。 | zh_TW |
dc.description.abstract | With the trend of industry 4.0 in recent years, how to apply Smart manufacturing into product lines has become the subject greatly focused in the industry. Nowadays, in the product line of wooden product processing, such as wood panel and wood veneer, most defect detection is still proceeded artificially. Product defect detection is one of the key part in the product lines. Thus, if we can introduce Automated Optical Inspection into product lines, this will be a great point to head for intellectualizing product lines. However, for the detected target like woods whose exteriors are complicated and nonregulated, it’s hard to find out the defects via traditional computer vision algorithm. With the rapid development of artificial intelligence and the technology of deep learning, many object detection model have been proposed. The kind of models has high adaptability and capabilities of finding targets from the messy background. To applying them as a measure of detecting defects possesses outstanding potential. Yet, object detection models needs a great amount of image training to have better detecting results. Within practical experiences, it’s quite common to face that not enough samples of products results in unideal detecting results. If we could overcome the problems, the possibility of applying the object detection models to the product lines will be elevated significantly. The research will practically apply the object detection model, evaluate whether it suits for the defects detection of wooden exteriors or not, and study the differences of detection effects on the different Feature Extractor. Besides, the research will use the image generating model to generate images of wooden defects, testing if training with image generating together with fewer original samples can elevate the defect detection effects on the object detection models. The research will eventually propose a best detecting set for wooden exterior defects according to the final result. Through the tests of the research, under the situation with enough samples, the training result of Faster R-CNN has almost 60 percent recall no matter we chose VGG16, Resnet50 or InceptionV2 as feature extractors. Under the situation with fewer samples, adding the generated defect images of Pix2Pix or GuaGan to the training will elevate the recall. For instance, training InceptionV2 as feature extractor with the defect images generated by GuaGan together can elevate the recall from 41.58% to 89.34%, which presented a significant effect on improving the results. Thus, using the set as a measure for the detection for wooden exterior defects showed a high potentiality. | en |
dc.description.provenance | Made available in DSpace on 2021-06-08T00:25:49Z (GMT). No. of bitstreams: 1 U0001-0402202118081300.pdf: 7446111 bytes, checksum: 6d57119d19c5fb810f28aa12d4ca2a93 (MD5) Previous issue date: 2021 | en |
dc.description.tableofcontents | 口試委員會審定書 I 誌謝 II 摘要 IV Abstract V 目錄 VII 圖目錄 IX 表目錄 XIII 第一章 緒論 1 1.1 研究背景 1 1.2 研究動機 4 1.3 研究目的 6 1.4 研究方法 7 1.5 論文架構 10 第二章 文獻回顧 11 2.1 深度學習於影像辨識 11 2.1.1 深度學習發展與應用 11 2.1.2 物件偵測網路 17 2.2 影像生成模型 21 2.3 數據增強於物件偵測 24 小結 25 第三章 深度學習模型 26 說明 26 3.1 區域卷積網絡 (Faster R-CNN) 26 3.2 特徵擷取器 29 3.3 條件生成網路 (Pix2Pix) 32 3.4 語義生成網路 (SPADE) 33 小結 35 第四章 實驗流程與方法 36 說明 36 4.1 資料集建立 37 4.1.1 光學取像 38 4.1.2 瑕疵影像擷取 42 4.1.3 瑕疵標記 44 4.1.4 製作訓練測試集 47 4.2 物件偵測模型建立 49 4.2.1 建立訓練環境 50 4.2.2 取得瑕疵資訊 53 4.2.3 訓練Faster R-CNN 模型 54 4.3 影像生成模型建立 61 4.3.1 訓練生成模型 61 4.3.2 生成瑕疵影像 65 4.4 應用生成模型於少樣本木材瑕疵檢測 67 4.4.1 傳統數據增強方法 67 4.4.2 實驗組合設計 68 4.5 木材瑕疵檢測訓練系統 71 小結 76 第五章 實驗結果與討論 77 說明 77 5.1 瑕疵檢出率 77 5.2 實驗結果 81 5.2.1 建立比較基準 81 5.2.2 加入訓練影像 82 5.2.3 結果討論 85 5.3 合併瑕疵類別實驗 93 小結 97 第六章 結論與未來展望 98 6.1 結論 98 6.2 未來展望 100 參考文獻 101 | |
dc.language.iso | zh-TW | |
dc.title | 應用影像生成模型於少樣本木材表面瑕疵檢測 | zh_TW |
dc.title | Application of Image Generating Models in Wood Surface Defect Detection with Small Sample Size | en |
dc.type | Thesis | |
dc.date.schoolyear | 109-1 | |
dc.description.degree | 碩士 | |
dc.contributor.oralexamcommittee | 李貫銘(Kuan-Ming Li),許智欽(Chih-Chin Hsu),林威延(Wei-Yen Lin) | |
dc.subject.keyword | 智慧製造,人工智慧,深度學習,物件偵測,影像生成,木材表面, | zh_TW |
dc.subject.keyword | Smart Manufacturing,Artificial Intelligence,Deep Learning,Object Detection,Image Generating,Wood Surface Defect, | en |
dc.relation.page | 103 | |
dc.identifier.doi | 10.6342/NTU202100532 | |
dc.rights.note | 未授權 | |
dc.date.accepted | 2021-02-05 | |
dc.contributor.author-college | 工學院 | zh_TW |
dc.contributor.author-dept | 機械工程學研究所 | zh_TW |
顯示於系所單位: | 機械工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
U0001-0402202118081300.pdf 目前未授權公開取用 | 7.27 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。