請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/94317完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 孫珍理 | zh_TW |
| dc.contributor.advisor | Chen-li Sun | en |
| dc.contributor.author | 馮其安 | zh_TW |
| dc.contributor.author | Chi-An Feng | en |
| dc.date.accessioned | 2024-08-15T16:47:12Z | - |
| dc.date.available | 2024-08-16 | - |
| dc.date.copyright | 2024-08-15 | - |
| dc.date.issued | 2024 | - |
| dc.date.submitted | 2024-08-05 | - |
| dc.identifier.citation | [1] U. Schmidt, M. Weigert, C. Broaddus, and G. Myers, "Cell detection with star-convex polygons," in Proceedings of. Medical Image Computing and Computer Assisted Intervention – MICCAI 2018, Cham, 2018, pp. 265-273: Springer International Publishing, doi: https://doi.org/10.48550/arXiv.1806.03535.
[2] Y. Kim and H. Park, "Deep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows," Scientific Reports, vol. 11, no. 1, p. 8940, 2021, doi: 10.1038/s41598-021-88334-0. [3] K. Lv, L. Zhou, Y. Chen, and H. Gao, "Experimental investigation on the evolution of bubble behavior in subcooled flow boiling in narrow rectangular channel based on bubble tracking algorithm," Frontiers in Energy Research, Original Research vol. 11, 2023, doi: 10.3389/fenrg.2023.1224306. [4] J. R. Hernandez-Aguilar and J. A. Finch, "Validation of bubble sizes obtained with incoherent imaging on a sloped viewing window," Chemical Engineering Science, vol. 60, no. 12, pp. 3323-3336, 2005, doi: https://doi.org/10.1016/j.ces.2004.12.022. [5] M. M. Hoque, S. Mitra, and G. Evans, "Bubble size distribution and turbulence characterization in a bubbly flow in the presence of surfactant," Experimental Thermal and Fluid Science, vol. 155, 2024, doi: https://doi.org/10.1016/j.expthermflusci.2024.111199. [6] K. A. Mat Said, A. Jambek, and N. Sulaiman, "A study of image processing using morphological opening and closing processes," International Journal of Control Theory and Applications, vol. 9, pp. 15-21, 2016. [7] K. He, G. Gkioxari, P. Dollár, and R. B. Girshick, "Mask R-CNN," 2017 IEEE International Conference on Computer Vision (ICCV), pp. 2980-2988, 2017, doi: https://doi.org/10.48550/arXiv.1703.06870. [8] K. O'shea and R. Nash, "An introduction to convolutional neural networks," arXiv preprint arXiv:1511.08458, 2015. [9] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You only look once: unified, real-time object detection," 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 779-788, 2015. [10] Y. Fu and Y. Liu, "BubGAN: Bubble generative adversarial networks for synthesizing realistic bubbly flow images," Chemical Engineering Science, vol. 204, pp. 35-47, 2019, doi: https://doi.org/10.1016/j.ces.2019.04.004. [11] H. Hessenkemper, S. Starke, Y. Atassi, T. Ziegenhein, and D. Lucas, "Bubble identification from images with machine learning methods," International Journal of Multiphase Flow, vol. 155, 2022, doi: https://doi.org/10.1016/j.ijmultiphaseflow.2022.104169. [12] D. Jha, M. A. Riegler, D. Johansen, P. Halvorsen, and H. D. Johansen, "DoubleU-Net: A Deep Convolutional Neural Network for Medical Image Segmentation," in Proceedings of. 2020 IEEE 33rd International Symposium on Computer-Based Medical Systems (CBMS), 2020, pp. 558-564, doi: 10.1109/CBMS49503.2020.00111. [13] K. He, X. Zhang, S. Ren, and J. Sun, "Deep residual learning for image recognition," in Proceedings of. Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770-778. [14] Y. Cui, C. Li, W. Zhang, X. Ning, X. Shi, J. Gao, and X. Lan, "A deep learning-based image processing method for bubble detection, segmentation, and shape reconstruction in high gas holdup sub-millimeter bubbly flows," Chemical Engineering Journal, vol. 449, 2022, doi: https://doi.org/10.1016/j.cej.2022.137859. [15] W. Burger, "Zhang’s camera calibration algorithm: in-depth tutorial and implementation," HGB16-05, pp. 1-6, 2016. [16] G. Bradski, "The OpenCV Library," Dr. Dobb's Journal of Software Tools, 2000. [17] H. Gholamalinejad and H. Khosravi, Pooling Methods in Deep Neural Networks, a Review. 2020. [18] O. Ronneberger, P. Fischer, and T. Brox, "U-Net: Convolutional Networks for Biomedical Image Segmentation," in Proceedings of. Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, Cham, 2015, pp. 234-241: Springer International Publishing. [19] A. Neubeck and L. V. Gool, "Efficient Non-Maximum Suppression," in Proceedings of. 18th International Conference on Pattern Recognition (ICPR'06), 2006, vol. 3, pp. 850-855, doi: 10.1109/ICPR.2006.479. [20] R. Bitter, T. Mohiuddin, and M. Nawrocki, LabVIEW: Advanced programming techniques. Crc Press. [21] H.-Y. Huo, 2022, "Enhancement of saturated pool boiling of HFE-7100 by multilayered copper meshes with porosity gradient," M.S. Thesis, National Taiwan University, Mechanical Engineering, Taipei. [22] A. Abd Al-salam Selami and A. Fadhil, "A Study of the Effects of Gaussian Noise on Image Features," Kirkuk University Journal / Scientific Studies (1992-0849), vol. 11, pp. 152-169, 2016, doi: 10.32894/kujss.2016.124648. [23] T.-Y. Lin, P. Dollár, R. Girshick, K. He, B. Hariharan, and S. Belongie, "Feature pyramid networks for object detection," in Proceedings of. Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 2117-2125. | - |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/94317 | - |
| dc.description.abstract | 本研究使用機器學習技術，對HFE-7100介電流體的池沸騰所產生的氣泡進行了深入分析。通過應用StarDist[1]模型對影像中的氣泡進行識別和形狀描繪，獲得了畫面中氣泡的大小分布和數量，並進一步進行氣泡追蹤，以分析氣泡的移除率。此外，我們在訓練資料中加入了區域陰影，以比較最終模型對氣泡識別能力的影響。研究中，我們利用提出的演算法對實驗所得的氣泡影像進行分析，同時我們使用本研究的模型對其他實驗所拍攝的氣泡影像進行分析，以驗證模型的泛用性。
實驗結果顯示，神經網路對於低熱通量條件下的單獨氣泡具有良好的識別效果，能夠準確分辨氣泡的邊緣和數量。在氣泡相鄰或互相遮擋時，演算法仍能正常辨識氣泡範圍並進行分割。在此階段，大部分氣泡為小型氣泡，且背景無大量氣泡遮擋產生的陰影，因此使用無陰影模型進行分析即可。隨著熱通量的提高，背景開始出現大量氣團和陰影，影像品質下降，且氣泡的大小增加，辨識難度加大。此時，使用包含陰影資料訓練的模型進行分析，可以觀察到氣泡辨識的正確性有顯著的提升。當池沸騰進入連續氣泡區域時，氣泡容易大量合併，導致氣泡面積急劇增加。儘管如此，模型仍能正常辨識大部分氣泡，但部分大型氣泡可能出現切割現象，導致體積計算產生誤差，進而低估氣泡體積。同時我們發現本模型亦可以應用於其研究之氣泡影像，在大部分情況下也表現良好，驗證了模型的泛用性。 | zh_TW |
| dc.description.abstract | In this study, machine learning techniques were used to analyze bubbles generated during pool boiling of the dielectric fluid HFE-7100. The StarDist[1] model was applied to identify and outline bubble shapes in images, allowing us to determine the bubble size distribution. Bubble tracking was also performed to analyze the removal rates of vapor. Shadow was added to the background of the training data to assess robustness of our model for bubble recognition under different conditions. In addition, both bubble images from experiments and other studies were tested to validate its generalizability.
The results showed that the neural network could effectively identify individual bubbles under low heat flux conditions, accurately distinguish bubble edges and counts. Even for adjacent or overlapping bubbles, recognition and segmentation were done successfully. At low heat flux, most bubbles and the effect of shadows were small, so the model trained by original data was good enough. As the heat flux increased, background shadows emerge and bubble grew, reducing the image quality. To address this issue, the model was re-trained by data manipulated with shadow-gradient background. This significantly improved the accuracy of bubble recognition under these conditions. When continuous bubble generation, led to coalescence, most bubbles could still be identified. However, some large bubbles were partially segmented, causing the underestimation of vapor volume. The model also performed well on bubble images from other studies, confirming its generalizability. | en |
| dc.description.provenance | Submitted by admin ntu (admin@lib.ntu.edu.tw) on 2024-08-15T16:47:12Z No. of bitstreams: 0 | en |
| dc.description.provenance | Made available in DSpace on 2024-08-15T16:47:12Z (GMT). No. of bitstreams: 0 | en |
| dc.description.tableofcontents | 口試委員會審定書 ii
謝誌 ii 摘要 iii Abstract iiv 目次 v 符號索引 viii 圖次 xiv 表次 xv 第一章導論 1 1.1 前言 1 1.2 文獻回顧 3 1.2.1 氣泡型態及其參數 3 1.2.2 影像語意分割方法 3 1.2.3 物件辨識方法 4 1.2.4 混和型態應用 4 1.3 研究目的 5 第二章 實驗架構與理論介紹 6 2.1 實驗架構 6 2.1.1 液體容器 6 2.1.2加熱模塊 7 2.1.3 冷卻器材 7 2.1.4 影像拍攝器材與擷取系統 8 2.1.5 影像處理設備 8 2.2 高速相機工作範圍量測與校正 9 2.2.1 相機參數校正理論 9 2.2.2 高速相機參數校正程序 11 2.3 氣泡辨識演算法理論 11 2.3.1 捲積神經網路 11 2.3.2 池化層 13 2.3.3 U-Net結構 14 2.3.4 非極大值抑制(Non-Maximum Suppression) 15 2.3.5 stardist理論 16 2.3.6 氣泡追蹤方法 18 2.3.7 神經網路訓練影像 19 2.3.8 氣泡體積還原方法 20 2.4 實驗與分析程序 20 2.4.1 實驗影像擷取 20 2.4.2 影像處理程序 21 2.5 誤差分析 23 第三章 實驗結果與討論 24 3.1 實驗細節與參數設置 24 3.1.1 訓練資料設定參數 24 3.1.2 模型參數設定 25 3.1.3 拍攝手法分析 27 3.2 氣泡辨識訓練結果 27 3.2.1 損失函數分析 27 3.2.2 模型輸出層之呈現 29 3.2.3 測試資料結果 30 3.2.4 實驗資料辨識結果 30 3.2.5 模型泛用性比較 33 3.3 訓練資料比較 34 3.4 氣泡變化的比較 36 3.4.1 氣泡面積變化 37 3.4.2 氣泡移除率比較 38 第四章 結論與建議 40 4.1 結論 40 4.2 建議 41 參考文獻 42 附錄 44 | - |
| dc.language.iso | zh_TW | - |
| dc.subject | 氣泡辨識 | zh_TW |
| dc.subject | 池沸騰 | zh_TW |
| dc.subject | HFE-7100 | zh_TW |
| dc.subject | 機器學習 | zh_TW |
| dc.subject | 兩相浸沒式冷卻 | zh_TW |
| dc.subject | machine learning | en |
| dc.subject | two phase immersion cooling | en |
| dc.subject | bubble detection | en |
| dc.subject | HFE-7100 | en |
| dc.subject | pool boiling | en |
| dc.title | 利用機器學習模型應用於池沸騰中氣泡之辨識分析 | zh_TW |
| dc.title | Machining learning enables analysis of bubble detection in pool boiling | en |
| dc.type | Thesis | - |
| dc.date.schoolyear | 112-2 | - |
| dc.description.degree | 碩士 | - |
| dc.contributor.oralexamcommittee | 廖英志;許麗;黃美嬌;柯正雯 | zh_TW |
| dc.contributor.oralexamcommittee | Ying-Chih Liao;Li Xu;Mei-Jiau Huang;Cheng-Wen Ko | en |
| dc.subject.keyword | 機器學習,氣泡辨識,兩相浸沒式冷卻,池沸騰,HFE-7100, | zh_TW |
| dc.subject.keyword | machine learning,bubble detection,two phase immersion cooling,pool boiling,HFE-7100, | en |
| dc.relation.page | 79 | - |
| dc.identifier.doi | 10.6342/NTU202403474 | - |
| dc.rights.note | 同意授權(限校園內公開) | - |
| dc.date.accepted | 2024-08-08 | - |
| dc.contributor.author-college | 工學院 | - |
| dc.contributor.author-dept | 機械工程學系 | - |
| 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 | |
|---|---|---|---|
| ntu-112-2.pdf 授權僅限NTU校內IP使用（校園外請利用VPN校外連線服務） | 4.75 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。