臺灣博碩士論文加值系統

由於工業製造對高精度、高自動化水平的要求，基於計算機視覺的尺寸檢測技術近年來備受關注。在大多數公司中，軸零件尺寸仍然是通過手動測量獲得的，滿足高速、高精度的測量要求很具有挑戰性。因此，開發一種低成本、省時、高精度、高效率的汽車傳動軸零件尺寸檢測系統至關重要。本論文提出了一種幾何軸類零件的尺寸檢測系統方法，此系統主要由兩個主要部分組成。首先，圖像採集系統用於計算機輔助設計進行設計和仿真，節省時間和金錢。第二部分，圖像預處理是通過二維離散傅里葉域方法中的自適應奇異值分解來消除光照變化對樣品表面的影響。提出了一種基於能量分析和大津閾值相結合的方法。 它不僅可以提取部分圖像的邊緣輪廓，而且具有無需參數調整、對噪聲敏感度低、各向同性等優點。汽車零件沿縱橫軸的產品尺寸是根據邊緣計算的。軸的大小是通過將軸中的像素數乘以一個像素的值來精確計算的。最後將汽車零件用於測量研究，得到測量數據並進行分析。實驗結果表明，該系統的重複性誤差小於0.3 mm。

Because of the high precision, and high level of automation requirements in industrial manufacturing, dimensional inspection technology based on computer vision has gained much attention in recent years. However, in most companies, dimensional shaft parts are still inspected by manual measurement. So, it is challenging to meet the measurement requirements of high speed and high precision. Therefore, it is vital to develop a dimensional inspection system for shaft parts that has a low cost, saves time, has high precision, and high efficiency. This dissertation proposes a dimensional inspection system approach for a geometric shaft part. The dimension inspection system mainly consists of two main parts. Firstly, the image acquisition system is utilized in computer-aided design for design and simulation saves time and money. In the second part, the image pre-processing is used by an adaptive singular value decomposition in the two-dimensional discrete Fourier domain method to eliminate the unexpected effect of lighting variation on the surface of the sample. A method based on a combination of energy analysis and Otsu's threshold is proposed. It not only extracts the edge contour of the image but also has the strengths of no parameter adjustment and low sensitivity to noise. The product size of the car part along the vertical and horizontal axes is calculated based on the edge. The size of the shaft is precisely calculated by multiplying the number of pixels in the shaft by the value of one pixel. Finally, the car part is used in the measurement study, and the measured data is obtained and analyzed. The experimental results show that the repeatability error of the system is less than 0.3 mm.

TABLE OF CONTENTS

LIST OF FIGURES xii
LIST OF TABLES……………………………………………………………………xiii
ABBREVIATIONS xiv
CHAPTER 1 INTRODUCTION 1
1.1 Motivation 1
1.2 Cart Part Overview 3
1.2.1 Heat Treatment 6
1.2.2 Geometric Dimensioning and Tolerancing 8
1.3 Dissertation Organization 11
1.4 Summary 11
CHAPTER 2 THE IMAGE ACQUISITION SYSTEM 12
2.1 Image acquisition system 12
2.1.1 Camera system 13
2.1.2 Lighting system 15
2.1.3 Fixture System 18
2.2 The Car Part Dataset 19
2.3 Summary 20
CHAPTER 3 CAR PART DIMENSIONAL INSPECTION SYSTEM 21
3.1 Image Enhancement for The Car Part 23
3.1.1 Singular Value Decomposition in Digital Image Processing 23
3.1.2 Fourier Transform in Digital Image Processing 24
3.1.3 Illumination Compensation for Car Part by Using Adaptive SVD in the Fourier Domain 27
3.2 Car Part Edge Detection 30
3.2.1 Energy analysis 33
3.2.2 Otsu’s Thresholding Method 34
3.3 Shaft Dimension Calculation 37
3.4 Summary 40
CHAPTER 4 EXPERIMENTAL RESULT AND DISCUSSION 41
4.1 Image Acquisition System Configuration 41
4.2 Car Part Inspection Experiment 43
4.3 Summary 47
CHAPTER 5 CONCLUSION AND FUTURE WORK 48
5.1 Conclusion 48
5.2 Future work 50
REFERENCES………... 51

[1] R. S. Khurmi and J. K. Gupta, "Shafts," in A textbook of Machine Design, 14th ed., S Chand & Co Ltd, July. 2005.
[2] T. Meshii, "Ferritic steel that ASTM E1921 master curve failed to characterize its fracture toughness temperature dependence,," Transactions of the JSME, pp. 7910-8507, 2019.
[3] J. Jeong, J. -H. Kim, S. -G. Choi, and Y. -T. Cho, "Mechanical Properties of White Metal on SCM440 Alloy Steelby Laser Cladding Treatment," applied sciences, vol. 11, 2021.
[4] T. G. Digges and S. J. Rosenberg, "Heat treatment of steel,," in Heat treatment and properties of Iron and Steel, Washington D.C, US. Government Printing Office, 1996.
[5] A. Rajan, T.V. Sharma and C.P. Sharma, "Heat treament processes for steels," in Heat treatment: Principles and Techniques, 2nd ed., PHI Learning Pvt. Ltd.,, 2011.
[6] A. Krulikowski, "GD& T Fundamentals,," in Fundamentals of geometric dimensioning and tolerancing, 3rd ed., Cengage Learning, 1994.
[7] J.Sarkar, "Structure of Computer-Aided Design Modeling," in Computer Aided Design, 1st ed., CRC Press, 2017.
[8] R. Lenk and C. Lenk, "Lighting systems," in Practical Lighting Design with LEDs, 2nd ed., iley Press, 2017.
[9] D. Dechow, "Machine vision illumination," in The Fundamentals of Machine vision, Fanuc America Corporation, 2019.
[10] D. B. Giap, T. N. Le, J.-W. Wang, and C.-N. Wang, "Wavelet Subband-Based Tensor for Smartphone Physical Button Inspection," IEEE Access, vol. 9, pp. 107399-107415, 2021.
[11] M. Moonen and B. D. Moor, "Matrix Singular Value Decomposition," in Singular Value Decomposition and its application, 1st ed., Elsevier Science, 1995, pp. 40- 45.
[12] K. R. Rao, D. N. Kim, and J. J. Hwang, "Discrete Fourier Transform," in Fast Fourier Transform - Algorithms and Applications, Springer Science & Business Media B.V, 2010.
[13] J. W. Wang, N. T. Le, J.- S. Lee, and C.- C. Wang, "Color face image enhancement using adaptive singular value decomposition in fourier domain for face recognition," Pattern Recognit., vol. 9, no. 2, pp. 31-49, Sept. 2016.
[14] E. R. Davies, "Edge Detection," in Machine Vision, Academic Press, 2005, pp. 200- 210.
[15] P. Ganesan, V. Rajini, B. S. Sathish, V. Kalist and S. K. K. Basha, "Satellite Image Segmentation Based on YCbCr Color Space," Indian Journal of Science and Technology, vol. 8, p. 35–41, 2015.
[16] N. T. Le, J.- W. Wang, C.-C.Wang and T. N. Nguyen, "Novel framework based on HOSVD for ski goggles defect detection and classification," Sensors, vol. 19, no. 24: 5538, pp. 10-15, 2019.
[17] D. M. Etter and J. A. Ingber, "Two-Dimensional Arrays," in Engineering Problem Solving with C++, 3rd ed., Pearson, pp. 60-75,2011.