農業影像二元分類：坵塊分離的檢測

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：13

、訪客IP：18.119.111.9

姓名

張淯淞(Yu-Sung Chang) 查詢紙本館藏

畢業系所

資訊工程學系

論文名稱

農業影像二元分類：坵塊分離的檢測
(Binary Classification of Agricultural Images: Detection of Parcel Separation)

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
★ 針織布異常偵測方法研究	★ 應用增量式學習於多種農作物判釋之研究
★ 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)

每年政府會使用含有農作物標記的坵塊向量圖（簡稱坵塊圖），對坵塊的資訊進行管理。但坵塊可能會隨著時間有所變化，有部份的坵塊會在新年度發生分離的情況，這時就需要對該坵塊所對應的坵塊向量進行修改。不過，出動人力查找坵塊分離是一件曠日費時的任務，因此使用 AI 分類器自動找出新年度航照圖中的坵塊分離儼然是一個重要的議題。過去坵塊變異的檢測研究多是使用影像相似度比對的方式來了解同一個坵塊是否有變化，而本研究則是提出一套產生 “仿坵塊分離使用情形” 合成資料的方法，與兩個判斷坵塊是否分離的影像二分類器。我們使用 “仿坵塊分離使用情形” 的合成資料訓練我們提出的影像二分類器，使其學會坵塊分離的影像樣態，並且使用訓練好的分類器來檢測真實航照中的坵塊分離。此外，本研究的貢獻在於透過實驗，我們驗證了真實測試資料集當中存在的幾種特殊的坵塊影像，會對我們所提出之影像二分類器造成判釋結果的影響。這個發現可提供後續研究在準備模型訓練資料集上的指引，未來的研究能遵循這個指引，針對特殊的坵塊影像作資料集的準備，並進一步探索判釋特殊坵塊影像的方法。

摘要(英)

Every year the government uses a parcel vector(aka polygon) shapefile with crop labels to manage the information of parcels. However since parcels may change over time,
some parcels will be separated in the new year, and the parcel vector corresponding to the parcel needs to be modified. Finding the parcel separations manually is a time-consuming task, so it is an important issue to use AI classifiers to automatically detect the separation of parcel in the new year’s aerial images. In the past, most researches on the detection of parcel change used image comparisons to check whether the same parcel changes. In
this paper, we propose a way to synthesize the ”Parcel separation pseudo data” and two binary classifier for predicting Parcel separation. We use the ”Parcel separation pseudo data” to train the several proposed binary classifiers, and use the trained classifier to
detect Parcel separation in real aerial image. Besides, the contribution of this study also include that there are some kinds of special parcel image in the real test dataset affect the accuracy of our proposed classifier. This finding can provide a valuable guidance for future research in preparing training datasets. Future research can follow this guideline to prepare datasets for special parcel images, and further explore the method of interpreting special parcel images.

關鍵字(中)

★ 坵塊變動檢測
★ 坵塊分類
★ VGG16-UNet
★ NetVLAD
★ CBAM

關鍵字(英)

★ Parcel Change Detection
★ Parcel Classification
★ VGG16-UNet
★ NetVLAD
★ CBAM

論文目次

摘要 vi
Abstract vii
誌謝 viii
目錄 ix
一、緒論 1
1.1 研究背景 ...................................................................... 1
1.2 研究動機與目的............................................................... 2
1.2.1 問題定義............................................................... 3
1.2.2 二分類規則 ............................................................ 3
1.3 研究貢獻 ...................................................................... 6
1.4 論文架構 ...................................................................... 6
二、相關研究 7
2.1 坵塊與坵塊圖 ................................................................. 7
2.2 坵塊變動偵測 ................................................................. 7
2.3 使用 VGG16-UNet 模型對航照影像判釋..................................... 12
2.4 全局平均池化 ................................................................. 12
2.5 NetVLAD ..................................................................... 13
2.6 Convolutional Block Attention Module....................................... 13
三、研究方法 15
3.1 資料前處理.................................................................... 15
3.1.1 坵塊向量合併.......................................................... 16
3.1.2 影像裁切............................................................... 18
ix
目錄
3.1.3 通道縮減............................................................... 19
3.1.4 影像增強............................................................... 20
3.2 影像分類模型 ................................................................. 21
3.2.1 VGG16-UNet-GAP 分類器............................................ 21
3.2.2 VGG16-UNet-NetVLAD 分類器 ...................................... 23
3.3 模型訓練方法 ................................................................. 23
四、實驗與結果討論 26
4.1 實驗資料集介紹............................................................... 26
4.2 評估方法 ...................................................................... 27
4.2.1 F1-Score ............................................................... 27
4.2.2 Two-tailed T-Test...................................................... 28
4.3 實驗一：內部測試 ............................................................ 30
4.3.1 實驗動機與目的 ....................................................... 30
4.3.2 實驗方法............................................................... 30
4.3.3 實驗結果............................................................... 35
4.4 實驗二：真實測試 ............................................................ 40
4.4.1 實驗動機與目的 ....................................................... 40
4.4.2 實驗方法............................................................... 40
4.4.3 實驗結果............................................................... 41
五、結論與未來展望 44
5.1 結論 ........................................................................... 44
5.2 未來展望 ...................................................................... 46
參考文獻 47

參考文獻

[1] 常健行. (民國 111 年). 111 年度農業地理資訊整合應用計畫（計畫編號：111 農科1.5.1-資-i3）. 行政院農業委員會, [Online]. Available: https://www.grb.gov.tw/search/planDetail?id=14581585.
[2] 施品妤, “結合語義分割與全連接網路做基於坵塊的水稻判釋之初步研究,” 碩士論文, 國立中央大學, 民國 111 年 1 月.
[3] J. Cohen, “A coefficient of agreement for nominal scales,” Educational and Psychological Measurement, vol. 20, no. 1, pp. 37–46, 1960.
[4] C. Van Rijsbergen, “Information retrieval,” 1979.
[5] 洪銘德. (民國 100 年 10 月). 台灣農地重劃辦理成果及省思. 行政院農業委員會農田水利處, [Online]. Available: https://www.coa.gov.tw/ws.php?id=24221.
[6] W. Bin, Y. Jian, Z. Zhongming, M. Yu, Y. Anzhi, C. Jingbo, H. Dongxu, L. Xingchun, and L. Shunxi, “Parcel-based change detection in land-use maps by adopting the holistic
feature,” IEEE J-STARS, vol. 7, no. 8, pp. 3482–3490, 2013.
[7] B. Wu, Y. Liu, J. Zhu, C. Lu, H. Yin, D. He, and A. Yue, “Research of optimal parameters for parcel-based change detection,” in IGARSS, Beijing, China, July 2016, pp. 2320–2323.
[8] O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in MICCAI, Munich, Germany, October 2015, pp. 234–241.
[9] I. Wahyuni, W.-J. Wang, D. Liang, and C.-C. Chang, “Rice semantic segmentation using unet-vgg16: A case study in yunlin, taiwan,” in ISPACS, Hualien, Taiwan, November
2021, pp. 1–2.
[10] K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” in ICLR, San Diego, USA, May 2015.
[11] M. Lin, Q. Chen, and S. Yan, “Network in network,” in ICLR, Banff, Canada, April 2014.
[12] 陳昭瑋, “應用卷積神經網路於航攝影像做基於坵塊的水稻判釋之研究,” 碩士論文, 國立中央大學, 民國 111 年 9 月.
[13] H. Jégou, M. Douze, C. Schmid, and P. Pérez, “Aggregating local descriptors into a
compact image representation,” in CVPRW, San Francisco, USA, June 2010, pp. 3304–3311.
[14] R. Arandjelovic, P. Gronat, A. Torii, T. Pajdla, and J. Sivic, “Netvlad: Cnn architecture for weakly supervised place recognition,” in CVPR, Las Vegas, USA, June 2016, pp. 5297–5307.
[15] S. Xu, T. Fang, D. Li, and S. Wang, “Object classification of aerial images with bag-ofvisual words,” GRSL, vol. 7, no. 2, pp. 366–370, 2009.47
[16] S. Woo, J. Park, J.-Y. Lee, and I. S. Kweon, “Cbam: Convolutional block attention module,” in ECCV, Munich, Germany, September 2018, pp. 3–19.
[17] Arcgis, Accessed June 15, 2023. [Online]. Available: https://www.esri.com/en- us/arcgis/products/arcgis-pro/overview.
[18] Gdal, Accessed June 15, 2023. [Online]. Available: https://gdal.org/.
[19] J. Hu, L. Shen, and G. Sun, “Squeeze-and-excitation networks,” in CVPR, Salt Lake City, USA, June 2018, pp. 7132–7141.
[20] O. Oktay, J. Schlemper, L. L. Folgoc, M. Lee, M. Heinrich, K. Misawa, K. Mori, S.McDonagh, N. Y. Hammerla, B. Kainz, et al., “Attention u-net: Learning where to look
for the pancreas,” in MIDL, Amsterdam, Netherlands, July 2018.
[21] R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, “Gradcam: Visual explanations from deep networks via gradient-based localization,” in ICCV,
Venice, Italy, October 2017, pp. 618–626.

指導教授

梁德容張欽圳(De-Ron Liang Chin-Chun Chang)

審核日期

2023-8-3

推文