跳到主要內容

臺灣博碩士論文加值系統

(44.200.194.255) 您好!臺灣時間:2024/07/23 04:50
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:陳宥霖
研究生(外文):CHEN, YOU-LIN
論文名稱:基於特徵提取用在IC缺陷辨識的模型
論文名稱(外文):A Model For IC Defect Identification Based on Feature Extraction
指導教授:李政道李政道引用關係
指導教授(外文):LEE, JENG-DAO
口試委員:李政道陳建璋陳世欣
口試委員(外文):LEE, JENG-DAOCHEN, CHIEN-CHANGCHEN, SHI-XIN
口試日期:2024-01-09
學位類別:碩士
校院名稱:國立虎尾科技大學
系所名稱:自動化工程系碩士班
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2024
畢業學年度:112
語文別:中文
論文頁數:40
中文關鍵詞:YOLOv4水平投影自動光學偵測積體電路元件焊點
外文關鍵詞:YOLOv4horizontal projectionAutomatic optical inspectionintegrated circuit componentsolder joint
相關次數:
  • 被引用被引用:0
  • 點閱點閱:43
  • 評分評分:
  • 下載下載:9
  • 收藏至我的研究室書目清單書目收藏:0
本研究主要探討自動化生產線中電路板(Printed Circuit Board, PCB)上的組件積體電路(Integrated Circuit, IC)的缺陷檢測,由於PCB上的組件越來越小型化且密度越來越高,加上IC的複雜性和多樣性,檢測和辨識電路上的組件(IC)的缺陷是一個重要的問題,本研究應用自動光學檢測系統(Automated Optical Inspection, AOI)提出了一種基於特徵提取的檢測模型,來對IC的缺陷進行辨識。本研究方法分為五個步驟:數據收集、IC定位、IC校正、特徵提取和缺陷辨識,數據收集分為兩個部分,分別是電路板和IC的數據。IC定位是利用YOLOv4來對電路板上的IC進行定位,將定位好的IC放大後並減少背景雜訊,方便進行缺陷特徵之提取,並減少背景雜訊。IC校正是使用水平投影校正的方法,分析圖片上各個角度之間的黑色像素變化,來判斷IC之水平傾斜角度進行校正。特徵提取是使用圖片註釋工具來對接腳缺陷的部分進行框選。缺陷辨識是使用YOLOv4模型,學習1500張圖片,訓練預測模型,最後使用額外的200張測試數據對模型進行測試,缺陷區域的可辨識率達到98.2%,再將這些由模型辨識出的缺陷跟真實的缺陷進行比較,辨識準確率達到96.2%。
The study aimed to discuss the defect detection of components integrated circuit (ICs) on printed circuit boards (PCBs) in automated production lines. Due to the increasing miniaturization and higher density of components on PCBs, along with the complexity and diversity of ICs, detecting and identifying defects in these electronic components is a critical issue. This research applies an automatic optical inspection system (AOI) to propose a defect recognition model based on feature extraction for ICs. The research methodology consists of five steps: data collection, IC localization, IC correction, feature extraction, and defect-recognition. Data collection involves two parts, circuit board data, and ICs data. IC localization uses YOLOv4 to locate ICs on the circuit board and enlarge the positioned IC and reduce background noise, magnifying them for easier defect feature extraction and reducing background noise. IC correction uses a horizontal projection correction method, analyzing variations in black pixels among different angles in the image to determine and correct the horizontal tilt angle of the IC. Feature extraction is achieved by using image annotation tools to outline the regions of pin defects. Defect recognition involves training a YOLOv4 model on 1500 images and then using an additional 200 test images to evaluate the model's performance. The defection region recognition rate reaches 98.2%, and after comparing the recognized defects with actual defects, the recognition accuracy reaches 96.2%.
中文摘要 …………………………………………………………… i
Abstract …………………………………………………………… ii
誌謝 …………………………………………………………… iii
目錄 …………………………………………………………… iv
表目錄 …………………………………………………………… v
圖目錄 …………………………………………………………… vi
符號說明 …………………………………………………………… viii
第一章 緒論……………………………………………………… 1
1.1 前言……………………………………………………… 1
1.2 研究動機與目的………………………………………… 1
1.3 文獻回顧………………………………………………… 1
1.4 論文架構………………………………………………… 3
第二章 校正模型與缺陷辨識原理………………………………4
2.1 圖片處理………………………………………………… 4
2.1.1 大津二值化……………………………………………… 4
2.1.2 水平投影………………………………………………… 6
2.2 YOLOv4………………………………………………… 7
2.3 CSPDarknet53…………………………………………… 9
第三章 實驗數據收集與缺陷辨識模型建立方法………………12
3.1 數據收集與拍攝環境…………………………………… 12
3.2 IC定位………………………………………………… 13
3.3 IC校正………………………………………………… 14
3.4 特徵提取………..……………………………………….. 15
3.5 缺陷辨識………..……………………………………….. 17
3.6 實驗設備………………………………………………… 19
第四章 實驗結果呈現…………………………………………… 20
4.1 圖片處理………………………………………………… 20
4.2 IC定位…………………………………...……………… 22
4.3 IC校正………………………………………………… 23
4.4 特徵提取………………………………………………… 29
4.5 缺陷辨識………………………………………………… 32
第五章 結論與未來展望………………………………………… 33
5.1 結論……………………………………………………… 33
5.2 未來展望………………………………………………… 33
參考文獻 …………………………………………………………… 34
Extended Abstract …………………………………………………………… 36


[1]H. Loh and M. Lu, “Printed circuit board inspection using image analysis,” in IEEE Transactions on Industry Applications, vol. 35, no. 2, pp. 426-432, March-April 1999.
[2]A. Caliskan and G. Gurkan, “Design and realization of an automatic optical inspection system for pcb solder joints,” 2021 International Conference on INnovations in Intelligent SysTems and Applications (INISTA), Kocaeli, Turkey, pp. 1-6, 2021.
[3]G. Acciani, G. Brunetti and G. Fornarelli, “Application of neural networks in optical inspection and classification of solder joints in surface mount technology,” in IEEE Transactions on Industrial Informatics, vol. 2, no. 3, pp. 200-209, Aug. 2006.
[4]H. Zhao, J. Cheng and J. Jin, “Ni vision based automatic optical inspection (AOI) for surface mount devices: devices and method,” 2009 International Conference on Applied Superconductivity and Electromagnetic Devices, Chengdu, China, pp. 356-360, 2009.
[5]W. Dai, A. Mujeeb, M. Erdt and A. Sourin, “Towards automatic optical inspection of soldering defects,” 2018 International Conference on Cyberworlds (CW), Singapore, pp. 375-382, 2018.
[6]J. Park, Y. Yoo, U. Kim, D. Lee and J. Kim, “D3pointnet: dual-level defect detection pointnet for solder paste printer in surface mount technology,” in IEEE Access, vol. 8, pp. 140310-140322, 2020.
[7]H. Wu, R. Lei and Y. Peng, “Pcbnet: a lightweight convolutional neural network for defect inspection in surface mount technology,” in IEEE Transactions on Instrumentation and Measurement, vol. 71, pp. 1-14, 2022.
[8]N. Saif, et al. “Automatic license plate recognition system for bangla license plates using convolutional neural network.” TENCON 2019-2019 IEEE Region 10 Conference (TENCON). IEEE, 2019.
[9]G. Fontana, N. Iacono, S. Negri, G. Papadia, “Design of a multi-mode hybrid micro-gripper for surface mount technology component assembly,” Micromachines (Basel), Jul 21 2023.
[10]F. Wu and X. Zhang, “Feature-extraction-based inspection algorithm for ic solder joints,” in IEEE Transactions on Components, Packaging and Manufacturing Technology, vol. 1, no. 5, pp. 689-694, May 2011.
[11] C. Chen, Y. Lin and J. Shen, “License plate recognition model for tilt correction based on convolutional neural network,” 2022 IEEE 4th Eurasia Conference on IOT, Communication and Engineering (ECICE), Yunlin, Taiwan, pp. 336-340, 2022.


[12] Y. Kessentini, et al. “A two-stage deep neural network for multi-norm license plate detection and recognition.” Expert systems with applications 136 (2019): 159-170,2019.
[13] W. Han, Z. Huang, A. kuerban, M. Yan and H. Fu, “A mask detection method for shoppers under the threat of covid-19 coronavirus,” 2020 International Conference on Computer Vision, Image and Deep Learning (CVIDL), Chongqing, China, pp. 442-447, 2020.
[14] W. Huang, J. Wen, W. Gan, L. Hu and B. Luo, “Neighborhood correlation enhancement network for pcb defect classification,” in IEEE Transactions on Instrumentation and Measurement, vol. 72, pp. 1-11, 2023.
[15] A. Bochkovskiy, C. Wang, and H. Liao, “Yolov4: optimal speed and accuracy of object detection.” arXiv preprint arXiv:2004.10934,2020.
[16] C. Wang, H. Liao, I. Yeh, Y. Wu, P. Chen, J. Hsieh, “Cspnet: a new backbone that can enhance learning capability of CNN,” Proceedings - 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW,P.1571-1580, 2020.
[17] J. Redmon and A. Farhadi, “Yolov3: an incremental improvement computer science,” arXiv: 1804.02767, 2018.
[18] (日)谷岡廣樹.康鑫著,深度學習入門教室,莊永裕譯,臉譜(城邦)文化出版社,台北,2019。
[19] C. Chang, H. Chen and C. Lin, “Detecting multiclass defects of printed circuit boards in the molded-interconnect-device manufacturing process using deep object detection networks,” 2022 IEEE International Conference on Industrial Engineering and Engineering Management (IEEM), Kuala Lumpur, Malaysia, pp. 1536-1540, 2022.
[20] S. Zhang, C. Qu, C. Ru, X. Wang and Z. Li, “Multi-objects recognition and self-explosion defect detection method for insulators based on lightweight ghostnet-yolov4 model deployed onboard uav,” in IEEE Access, vol. 11, pp. 39713-39725, 2023.
[21] B. Luo, H. Wang, J. Jia, N. Qin, Y. Du and L. Xie, “Defect detection of metal sheets based on improved yolox algorithm,” 2023 CAA Symposium on Fault Detection, Supervision and Safety for Technical Processes (SAFE-PROCESS), Yibin, China, pp. 1-6, 2023.
[22] B. Wan et al., “Lfrnet: localizing, focus, and refinement network for sa lient object detection of surface defects,” in IEEE Transactions on Instrumenta-tion and Measurement, vol. 72, pp. 1-12, 2023.

QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top