跳到主要內容

臺灣博碩士論文加值系統

(44.192.94.177) 您好!臺灣時間:2024/07/21 18:36
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:阮文松
研究生(外文):VAN-TUNG NGUYEN
論文名稱:探討光彈特性於多波長與人工智慧去雜訊之應力分佈量化方法
論文名稱(外文):Quantification Approach of Stress Distribution by Photoelasticity with Multiple Wavelengths and AI Denoising
指導教授:連啓翔
指導教授(外文):CHI-HSIANG LIEN
口試委員:連啓翔潘國興張家源
口試委員(外文):CHI-HSIANG LIENQUOC-HUNG PHANCHIA-YUAN CHANG
口試日期:2023-11-23
學位類別:碩士
校院名稱:國立聯合大學
系所名稱:機械工程學系碩士班
學門:工程學門
學類:機械工程學類
論文種類:學術論文
論文出版年:2023
畢業學年度:112
語文別:英文
論文頁數:79
中文關鍵詞:AI 降噪多波長光測彈性力學偏振相機全場應力
外文關鍵詞:AI denoisingmultiple wavelengthsphotoelasticitypolarization camerawhole-stress
相關次數:
  • 被引用被引用:0
  • 點閱點閱:26
  • 評分評分:
  • 下載下載:2
  • 收藏至我的研究室書目清單書目收藏:0
本研究利用一個四向偏振相機(偏振方向為0°、45°、90°和135°)提出一種有效的應力分析方法之探討;其中以圓盤形狀之物件作為標準試片,測試與分析討論皆以圓偏振狀態入射光穿透試片產生之變化為例,而進行模擬與實驗。本研究分析五個不同波長對光彈性行為的影響,包含藍光(450nm)、綠光(543nm)、黃光(577nm)、橘光(612nm)和紅光(633nm)。為提高準確性,以633nm紅光波長進行人工智慧去噪的開發,其獲得之結果相較於傳統方式更準確。本研究中修改四個模型網絡(U-Net、VGG16、ResNet18和MobileNet-V2)進行實驗。結果顯示U-Net和VGG16獲得最高的準確性;若加入考量訓練成本ResNet18和MobileNet-V2可在準確性和運算成本兩者之間取得較好的條件。此外,使用AI去噪時,SSIM值最高為94.37%,NRMSE值最低為0.55%。本研究證明其在整個場地應力分析中的準確性,提高實際應力評估的效果,針對未來的複雜結構分析能夠開啟新的途徑方法之一。
This study presents a method for efficient stress analysis, focusing on a circular disc. The experiment used a 4-linear polarization state (0°, 45°, 90°, and 135°) and circular polarization state to capture images. The optimization of the setup allows elimination of additional analyzers including polarizer and quarter-wave plates after the specimen. A phase unwrapping strategy is established for both iso-chromatic and iso-clinic analyses. This method allows for the precise determination of whole-field stress, including normal stress, shear stress, and different principal stresses. Additionally, this study analysis the influence of five distinct wavelengths on photoelasticity behavior, namely blue (450nm), green (543nm), yellow (577), orange (612nm), and red (633nm). To enhance accuracy, AI denoising is applied with the red wavelength of 633nm. The results obtained are more accurate than manual filtering. The modification of four model networks (U-Net, VGG16, ResNet18, and MobileNet-V2) reveals effective results. U-Net and VGG16 achieve the highest accuracy results, while ResNet18 and MobileNet-V2 achieve a balance between accuracy and hardware cost. SSIM value is highest with 94.37%, and NRMSE values are lowest with 0.55% when using AI denoising. The study proves its accuracy for whole-field stress analysis, enhancing practical stress assessment. It opens avenues for intricate structural analyses in the future.  
Acknowledgment
摘要
Abstract
Contents
List of figures
List of tables
Chapter 1 Introduction
1.1 Overview
1.2 Motivation
1.3 Contributions
1.4 Structure of the thesis
Chapter 2 Methodology
2.1 Stress-optic law
2.2 Stress field in radially compressed circular disc
2.3 Image processing
2.3.1 Histogram
2.3.2 Median filter
2.4 Gaussian kernel regression
2.5 Iso-clinic and iso-chromatic pattern
2.6 Phase unwrapping methods
2.6.1 Unwrapping iso-chromatic
2.6.2 Unwrapping iso-clinic
2.7 Whole-field stress determination
2.8 Stress error evaluation
2.9 U-Net model
2.10 VGG16 modified with decoder block
2.11 ResNet18 modified with decoder block
2.12 MobileNet-V2 modified with decoder block
2.13 Training dataset and parameters
2.13.1 Training dataset
2.13.2 Data augmentation
2.13.3 Data training
2.14 AI denoising evaluation metrics
Chapter 3 Experiment and results
3.1 Whole-field stress determination strategy
3.2 Experiment model
3.3 Training results
3.3.1 Training loss and RMSE value
3.3.2 Peak signal to noise ratio result
3.3.3 Structural similarity index result
3.4 Four polarizations step image with red wavelength of 633nm
3.5 Phase map with multiple wavelength image
3.6 Image filters with red wavelength of 633nm
3.7 Phase unwrapping image
3.8 Fringe constant of material
3.9 Stress distribution and evaluation
3.9.1 Stress distribution using manual filter with red wavelength of 633nm
3.9.2 Stress evaluation with five-wavelengths 53
3.9.3 Stress evaluation by AI denoising with red wavelength of 633nm
3.9.4 Average stress evaluation using AI denoising and multiple wavelengths
Chapter 4 Conclusions
References

[1]Su, F., Lan, T., & Pan, X. (2015). Stress evaluation of Through-Silicon Vias using micro-infrared photoelasticity and finite element analysis. Optics and Lasers in Engineering, 74, 87–93.
[2]Liu, W., Ma, Z., Li, L., & Yue, Z. (2020). Photoelastic evaluation of stress fields and notch stress intensity factors for blunt V-notches. Theoretical and Applied Fracture Mechanics, 110, 102806.
[3]Aben, H., Ainola, L., & Anton, J. (2000). Integrated photoelasticity for nondestructive residual stress measurement in glass. Optics and Lasers in Engineering, 33(1), 49–64.
[4]Asundi, A., & Kishen, A. (2000). A strain gauge and photoelastic analysis of in vivo strain and in vitro stress distribution in human dental supporting structures. Archives of Oral Biology, 45(7), 543–550.
[5]Ramesh, K., & Sasikumar, S. (2020). Digital photoelasticity: Recent developments and diverse applications. Optics and Lasers in Engineering, 135, 106186.
[6]White, C. M., Smith, D. H., Jones, K. L., Goodman, A., Jikich, S. A., LaCount, R. B., DuBose, S. B., Ozdemir, E., Morsi, B. I., & Schroeder, K. (2005). Sequestration of Carbon Dioxide in Coal with Enhanced Coalbed Methane RecoveryA Review. Energy & Fuels, 19(3), 659–724.
[7]Ju, W. (2014). On area-specific underground research laboratory for geological disposal of high-level radioactive waste in China. Journal of Rock Mechanics and Geotechnical Engineering, 6(2), 99–104.
[8]Shaik, A. R., Rahman, S. S., Tran, N. H., & Tran, T. (2011). Numerical simulation of Fluid-Rock coupling heat transfer in naturally fractured geothermal system. Applied Thermal Engineering, 31(10), 1600–1606.
[9]Kim, S. O., Kim, J. J., Yun, S. T., & Kim, K. W. (2003). Numerical and Experimental Studies on Cadmium (II) Transport in Kaolinite Clay under Electrical Fields. Water, Air, & Soil Pollution, 150(1/4), 135–162.
[10]Brewster, D. C. (1816). On the communication of the structure of doubly refracting crystals to glass, muriate of soda, fluor spar, and other substances, by mechanical compression and dilatation. Philosophical Transactions of the Royal Society of London, 106, 156–178.
[11]Liu, X., & Dai, S. (2015). Cubic polynomial curve-guided method for isochromatic determination in three-fringe photoelasticity. Chinese Optics Letters, 13(10), 101202–101206.
[12]Theocaris, P. S., & Gdoutos, E. E. (1979). Matrix Theory of Photoelasticity. In SPringer series in optical sciences.
[13]Coker, E.G., & Filon, L.N. (1931). A treatise on photo-elasticity. CiNii Books.
[14]Oppel, G. (1937, April 1). The photoelastic investigation of three-dimensional stress and strain conditions. NASA Technical Reports Server (NTRS).
[15]Weller, R., & Bussey, J. K. (1940). Photoelastic Analysis of Three-Dimensional Stress Systems using Scattered Light. The Journal of the Royal Aeronautical Society, 44(349), 74–88.
[16]Aben, H., & Guillemet, C. (1993). Integrated Photoelasticity. In Springer eBooks (pp. 86–101).
[17]Ju, Y., Ren, Z., Li, W., Mao, L., & Chiang, F. (2018). Photoelastic method to quantitatively visualise the evolution of whole-field stress in 3D printed models subject to continuous loading processes. Optics and Lasers in Engineering, 100, 248–258.
[18]Xu, Z., Zhang, S., Han, Y., Dong, X., Su, Z., & Zhang, D. (2022). Full-field phase shifting and stress quantification using a polarization camera. Measurement, 201, 111727.
[19]Voloshin, A., & Bürger, C. (1983). Half-fringe photoelasticity: A new approach to whole-field stress analysis. Experimental Mechanics, 23(3), 304–313.
[20]Chen, T. Y. F. (1997). Digital determination of photoelastic birefringence using two wavelengths. Experimental Mechanics, 37(3), 232–236.
[21]Plouzennec, N., & Lagarde, A. (1999). Two-wavelength method for full-field automated photoelasticity. Experimental Mechanics, 39(4), 274–277.
[22]Kihara, T. (2003). An arctangent unwrapping technique of photoelasticity using linearly polarized light at three wavelengths. Strain, 39(2), 65–71.
[23]Hecker, F. W., & Morche, B. (1986). Computer-Aided measurement of relative retardations in plane photoelasticity. In Springer eBooks, 535–542.
[24]Yoneyama, S., Morimoto, Y., & Matsui, R. (2003). Photoelastic fringe pattern analysis by real-time phase-shifting method. Optics and Lasers in Engineering, 39(1), 1–13.
[25]Zhang, D., Han, Y., Bao, Z., & Arola, D. (2007). Automatic determination of parameters in photoelasticity. Optics and Lasers in Engineering, 45(8), 860–867.
[26]Ramji, M., & Ramesh, K. (2010). Adaptive quality guided phase unwrapping algorithm for Whole-Field digital photoelastic parameter estimation of complex models. Strain, 46(2), 184–194.
[27]Ajovalasit, A., Pitarresi, G., & Zuccarello, B. (2007). Limitation of carrier fringe methods in digital photoelasticity. Optics and Lasers in Engineering, 45(5), 631–636.
[28]Zuccarello, B. (2005). Complete isochromatic fringe-order analysis in digital photoelasticity by Fourier Transform and Load stepping. Strain, 41(2), 49–58.
[29]Ajovalasit, A., Barone, S., & Petrucci, G. (1995). Towards RGB photoelasticity: Full-field automated photoelasticity in white light. Experimental Mechanics, 35(3), 193–200.
[30]Madhu, K. R., Prasath, R. G. R., & Ramesh, K. (2007). Colour adaptation in three fringe photoelasticity. Experimental Mechanics, 47(2), 271–276.
[31]Ekman, M. J., & Nurse, A. D. (1998). Absolute determination of the isochromatic parameter by load-stepping photoelasticity. Experimental Mechanics, 38(3), 189–195.
[32]Ramesh, K., & Tamrakar, D. K. (2000). Improved determination of retardation in digital photoelasticity by load stepping. Optics and Lasers in Engineering, 33(6), 387–400.
[33]Shang, W., Ji, X., & Yang, X. (2015). Study on several problems of automatic full-field isoclinic parameter measurement by digital phase shifting photoelasticity. Optik, 126(19), 1981–1985.
[34]Xu, Z., Zhang, S., Han, Y., Dong, X., Su, Z., & Zhang, D. (2022). Full-field phase shifting and stress quantification using a polarization camera. Measurement, 201, 111727.
[35]Patterson, E. A., & Wang, Z. F. (1991). Towards full field automated photoelastic analysis of complex components. Strain, 27(2), 49–53.
[36]Kothiyal, M. P., & Delisle, C. (1985). Rotating analyzer heterodyne interferometer: error analysis. Applied Optics, 24(15), 2288.
[37]Wyant, J. C. (2003). Dynamic interferometry. Optics & Photonics News, 14(4), 36.
[38]Yoneyama, S. (2006). Simultaneous observation of phase-stepped photoelastic fringes using a pixelated microretarder array. Optical Engineering, 45(8), 083604.
[39]Yoneyama, S., & Arikawa, S. (2016). Instantaneous phase-stepping interferometry based on a pixelated micro-polarizer array. Theoretical and Applied Mechanics Letters, 6(4), 162–166.
[40]Siegmann, P., Bäckman, D., & Patterson, E. A. (2005). A robust approach to demodulating and unwrapping phase-stepped photoelastic data. Experimental Mechanics, 45(3), 278–289.
[41]Pinit, P., & Umezaki, E. (2007). Digitally whole-field analysis of isoclinic parameter in photoelasticity by four-step color phase-shifting technique. Optics and Lasers in Engineering, 45(7), 795–807.
[42]Ramji, M., & Ramesh, K. (2008). Whole field evaluation of stress components in digital photoelasticity—Issues, implementation and application. Optics and Lasers in Engineering, 46(3), 257–271.
[43]Ramji, M., & Ramesh, K. (2010). Adaptive quality guided phase unwrapping algorithm for Whole-Field digital photoelastic parameter estimation of complex models. Strain, 46(2), 184–194.
[44]Kasimayan, T., & Ramesh, K. (2009). Adaptive smoothing for IsoClinic parameter evaluation in digital photoelasticity. Strain.
[45]Wang, Y., Wu, G., Chen, G., & Chai, T. (2014). Data mining based noise diagnosis and fuzzy filter design for image processing. Computers & Electrical Engineering, 40(7), 2038–2049.
[46]Tojo, L., Maik, V., & Devi, M. (2022). Image denoising using multi scaling aided double decker convolutional neural network. Optik, 170350.
[47]Ramesh, K., Desai, S., Jariwala, D., & Shukla, V. (2022). AI Modelled Clutch Operation For Automobiles. 2022 IEEE World Conference on Applied Intelligence and Computing (AIC).
[48]Kumar, A., Finley, B., Braud, T., Tarkoma, S., & Hui, P. (2021). Sketching an AI Marketplace: tech, economic, and regulatory aspects. IEEE Access, 9, 13761–13774.
[49]Tsai, F. T., Nguyen, V. T., Duong, T. P., Phan, Q. H., & Lien, C. H. (2023). Tomato Fruit Detection Using Modified Yolov5m Model with Convolutional Neural Networks. Plants, 12(17), 3067.
[50]Phan, Q. H., Nguyen, V. T., Lien, C. H., Duong, T. P., Hou, M. T. K., & Le, N. B. (2023). Classification of tomato fruit using YoLov5 and convolutional neural network models. Plants, 12(4), 790.
[51]Furman, J., & Seamans, R. (2019). AI and the Economy. Innovation Policy and the Economy, 19, 161–191.
[52]Reader, A. J., & Pan, B. (2023). AI for PET image reconstruction. British Journal of Radiology, 96(1150).
[53]Rabbat, N., Qureshi, A. F., Hsu, K. T., Asif, Z., Chitnis, P. V., Shobeiri, S. A., & Wei, Q. (2023). Automated Segmentation of Levator Ani Muscle from 3D Endovaginal Ultrasound Images. Bioengineering, 10(8), 894.
[54]Zou, Y., Amidi, E., Luo, H., & Zhu, Q. (2022). Ultrasound-enhanced Unet model for quantitative photoacoustic tomography of ovarian lesions. Photoacoustics, 28, 100420.
[55]Ramesh, K. (2000). Digital Photoelasticity: Advanced Techniques and Applications. In Springer eBooks, Chapter 1, 1-42.
[56]Ju, Y., Ren, Z., Li, W., Mao, L., & Chiang, F. (2018). Photoelastic method to quantitatively visualise the evolution of whole-field stress in 3D printed models subject to continuous loading processes. Optics and Lasers in Engineering, 100, 248–258.
[57]Kong, N. S. P. (2013). A Literature review on histogram Equalization and its Variations for Digital Image Enhancement. International Journal of Innovation, Management and Technology.
[58]Mustafa, W. A., & Kader, M. M. M. A. (2018). A review of histogram Equalization Techniques in Image Enhancement Application. Journal of Physics, 1019, 012026.
[59]Agrawal, S., Panda, R., Mishro, P. K., & Abraham, A. (2022). A novel joint histogram equalization based image contrast enhancement. Journal of King Saud University - Computer and Information Sciences, 34(4), 1172–1182.
[60]Pitas, I., & Venetsanopoulos, A. N. (1990). Median filters. In Springer eBooks, 63–116.
[61]Wang, W., Xu, Z., Lu, W. H., & Zhang, X. (2003). Determination of the spread parameter in the Gaussian kernel for classification and regression. Neurocomputing, 55(3–4), 643–663.
[62]Zhang, H., & Brandt, T. D. (2021). Cleaning Images with Gaussian Process Regression. The Astronomical Journal, 162(4), 139.
[63]Nadaraya, É. A. (1964). On estimating regression. Theory of Probability and Its Applications, 9(1), 141–142.
[64]Su, F., & Wang, Z. (2023). Error analysis and correction of a photoelastic method based on a pixelated polarization camera. Optics and Lasers in Engineering, 161, 107374.
[65]Ramesh, K. (2000). Digital Photoelasticity: Advanced Techniques and Applications. In Springer eBooks, Chapter 9, 303-344.
[66]Guo, E., Liu, Y., Han, Y., Arola, D., & Zhang, D. (2018). Full-field stress determination in photoelasticity with phase shifting technique. Measurement Science and Technology, 29(4), 045208.
[67]Mukashev, D., Zhuzbay, N., Koshkinbayeva, A., Orazbayev, B., & Kappassov, Z. (2022). PhotoElasticFinger: Robot tactile fingertip based on photoelastic effect. Sensors, 22(18), 6807.
[68]Channappayya, S. S., Bovik, A. C., & Heath, R. W. (2008). Rate bounds on SSIM Index of quantized images. IEEE Transactions on Image Processing, 17(9), 1624–1639.
[69]Ronneberger, O., Fischer, P., & Brox, T. (2015). U-NET: Convolutional Networks for Biomedical Image Segmentation. In Lecture Notes in Computer Science, 234–241.
[70]Liu, S., & Deng, W. (2015). Very deep convolutional neural network based image classification using small training sample size. 2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR), 730–734.
[71]He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770–778.
[72]Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L. C. (2018). MobileNetV2: Inverted Residuals and Linear Bottlenecks. IEEE/CVF Conference on Computer Vision and Pattern Recognition, 4510–4520.
[73]Ouahabi, A. (2013). A review of wavelet denoising in medical imaging. 2013 8th International Workshop on Systems, Signal Processing and Their Applications (WoSSPA), 19–26.

QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top