針織布異常偵測方法研究

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：18

、訪客IP：3.143.9.115

姓名

林之詠(Chih-Yung Lin) 查詢紙本館藏

畢業系所

資訊工程學系

論文名稱

針織布異常偵測方法研究
(Knitted Fabric Anomaly Detection Method Research)

相關論文

★ 基於OpenPose特徵的行人分心偵測	★ 建構深度學習CNN模型以正確分類傳統AOI模型之偵測結果
★ 一種結合循序向後選擇法與回歸樹分析的瑕疵肇因關鍵因子擷取方法與系統－以紡織製程為例	★ 融合生成對抗網路及領域知識的分層式影像擴增
★ 應用增量式學習於多種農作物判釋之研究	★ 農業影像二元分類：坵塊分離的檢測
★ 應用遷移學習於胚布瑕疵檢測	★ 基於網頁 MCES 分析系統的分析過程加速
★ 使用一台智能手錶在有限的數據下檢測方向盤位置	★ 在基於操作行為之智慧型手機身分驗證系統上的使用者偽造攻擊研究
★ 移除弱特徵以抵禦基於行為身份驗證系統上的偽造攻擊

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 ( 永不開放)

摘要(中)

近年隨著機器學習、電腦視覺以及深度學習領域日新月異，提升自動光學檢測的效能與效率，各種產業紛紛嘗試使用自動光學檢測系統取代檢測員來進行產品上的瑕疵偵測。本研究主要目的為協助針織布廠商處理在針織布上進行自動光學檢測的過程中遭遇的困境，我們結合深度學習模型與經驗法則運算，設計一套能正確分析針織布影像上是否存在瑕疵之方法，解決廠商希望引進新檢測技術卻無法順利執行之問題，以取代傳統的肉眼檢測方式。
本研究設計從現場廠域中即時蒐集影像資料，並對影像進行標記，作為瑕疵偵測方法之訓練資料集。我們提出一套剖布線取樣方法，透過影像分割模型以及經驗法則運算，分析針織布中良品資料的特徵，並執行剖布線抽樣作業，進行影像挑選並製作成訓練資料集。接著我們提出另外一套瑕疵偵測方法，透過影像分割模型以及門檻值的設定，將每一張影像依照紋理進行判別，最後再分析連續影像之間的時序性關聯，執行投票運算，計算出輸入的連續影像是否為瑕疵品，作為系統輸出。後續能將系統偵測的瑕疵結果輸入到機台上的警報裝置，即時通知現場作業人員，維護針織布機台。

摘要(英)

Since machine learning, computer vision and deep learning have been developed quickly these years, the efficiency and effectiveness of AOI have been improved a lot. Recently, inspection in many industries has been replaced inspectors with AOI system. The main purpose of our research is trying to deal with problems which knitted fabric manufacturers encountered when deploying AOI system for inspection. We have developed a system which is capable to detect defects on knitted fabric correctly with combination of deep learning models and rule-based calculation, so that the manufactures are able to do inspection with AOI system rather than inspectors.
We plan to collect and label data on spot. After processing, those data will become training data of defect detection method. We proposed a “Cutline Sampling Method” to sample Cutline images by analyzing features with semantic segmentation model and rule-based calculation. We also proposed a “Rule-based Defect Detection Method”. Classification of each image will be
done by semantic segmentation model and thresholds. Every single image classified result will be sent to time series voting. According to the relation between adjacent images, the system will distinguish the consecutive input images between “Good Product” and “Defective Product”
as output. The output could be transmitted to alert system on knitted fabric weaving machine and inform workers so that they are able to maintain the weaving machine immediately.

關鍵字(中)

★ 異常偵測
★ 計算機視覺
★ 自動光學檢測
★ 影像分割模型

關鍵字(英)

★ Anomaly detection
★ Computer vision
★ AOI
★ Semantic segmentation model

論文目次

中文摘要 i
Abstract ii
目錄 iii
圖目錄 v
表目錄 vii
一、緒論 1
1-1 研究背景 1
1-2 研究動機 4
1-3 研究目的 8
1-4 研究貢獻 8
1-5 論文架構 9
二、相關研究 10
2-1 布料異常偵測 10
2-1-1 Entropy Filtering 10
2-1-2 MSCDAE 12
2-2 U-Net 16
2-3 Morphological Closing 18
三、系統架構 20
3-1影像輸入 22
3-2 剖布線取樣方法 24
3-2-1 語意分割模型 (剖布線取樣方法) 25
3-2-2 Morphological Closing 27
3-2-3 時序性特徵運算 27
3-2-4 剖布線抽樣 29
3-2-5 正常紋理抽樣 31
3-2-6 訓練資料集生成 32
3-3 瑕疵偵測方法 34
3-3-1 語意分割模型 (瑕疵偵測方法) 34
3-3-2 靜態分類 35
3-3-3 時序性投票 39
3-4 系統輸出 41
四、實驗與討論 42
4-1 實驗一 : 剖布線取樣方法之評估 42
4-1-1 問題定義 42
4-1-2 資料集 42
4-1-3 實驗方法 43
4-1-4 實驗結果 45
4-2 實驗二 : 瑕疵偵測方法之評估 48
4-2-1 問題定義 48
4-2-2 資料集 48
4-2-3 實驗方法 50
4-2-4 實驗結果 52
五、結論與未來研究 54
5-1 論文總結 54
5-2 未來研究 55
參考文獻 56

參考文獻

[1] P.Malhotra, L.Vig, G.Shroff, andP.Agarwal, “Long Short Term Memory networks for anomaly detection in time series,” 23rd Eur. Symp. Artif. Neural Networks, Comput.
Intell. Mach. Learn. ESANN 2015 - Proc., no. April, pp. 89–94, 2015.
[2] S.Ji, W.Xu, M.Yang, andK.Yu, “3D Convolutional neural networks for human action recognition,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 35, no. 1, pp. 221–231,
2013, doi: 10.1109/TPAMI.2012.59.
[3] M. S.Sayed, “Robust fabric defect detection algorithm using entropy filtering and minimum error thresholding,” Midwest Symp. Circuits Syst., vol. 0, no. October 2016,
2016, doi: 10.1109/MWSCAS.2016.7870031.
[4] L.Weninger, M.Kopaczka, andD.Merhof, “Defect detection in plain weave fabrics by yarn tracking and fully convolutional networks,” I2MTC 2018 - 2018 IEEE Int. Instrum. Meas. Technol. Conf. Discov. New Horizons Instrum. Meas. Proc., pp. 1–6, 2018, doi: 10.1109/I2MTC.2018.8409546.
[5] S.Mei, Y.Wang, andG.Wen, “Automatic fabric defect detection with a multi-scale convolutional denoising autoencoder network model,” Sensors (Switzerland), vol. 18, no. 4, pp. 1–18, 2018, doi: 10.3390/s18041064.
[6] W.Weng andX.Zhu, “INet: Convolutional Networks for Biomedical Image Segmentation,” IEEE Access, vol. 9, pp. 16591–16603, 2021, doi: 10.1109/ACCESS.2021.3053408.
[7] Z.Zhou, M. M.Rahman Siddiquee, N.Tajbakhsh, andJ.Liang, “Unet++: A nested u-net architecture for medical image segmentation,” Lect. Notes Comput. Sci. (including Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics), vol. 11045 LNCS, pp. 3–11, 2018, doi: 10.1007/978-3-030-00889-5_1.57
[8] H.Huang et al., “UNet 3+: A Full-Scale Connected UNet for Medical Image Segmentation,” ICASSP, IEEE Int. Conf. Acoust. Speech Signal Process. - Proc., vol. 2020-May, no. ii, pp. 1055–1059, 2020, doi: 10.1109/ICASSP40776.2020.9053405.
[9] J.Jing, Z.Wang, M.Rätsch, andH.Zhang, “Mobile-Unet: An efficient convolutional neural network for fabric defect detection,” Text. Res. J., no. May, 2020, doi:
10.1177/0040517520928604.
[10] R.Srisha andA.Khan, “Morphological Operations for Image Processing : Understanding and its Applications,” NCVSComs-13, no. December, pp. 17–19, 2013.
[11] X.Tao, D.Zhang, W.Ma, X.Liu, andDeXu, “Automatic metallic surface defect detection and recognition with convolutional neural networks,” Appl. Sci., vol. 8, no. 9, pp. 1–15, 2018, doi: 10.3390/app8091575.

指導教授

梁德容張欽圳

審核日期

2021-7-15

推文