未來展望

本研究針對植被指數的比較，對於日後需要挑選植被指數進行應用的研究，具 有一定程度的參考價值。同時，本研究提出之能自動決定斑點偵測所用參數數值的 流程，可在日後研究上，應用於當影像中欲偵測之目標，其大小、形狀皆非常相近 時的狀況。然而，本研究於秧苗偵測中所使用的參數，仍有部分為手動設定，若能 設計一套系統化的流程，自動決定其他參數的數值，即能提高秧苗偵測階段的自動 化程度。而研究中目前已做到單一時間點的秧苗對齊，未來則希望利用空拍圖與鑲 嵌圖，實現跨時間的秧苗對齊，以此掌握作物的成長幅度。此外，研究中所提出之 影像拼接方法仍然屬於線性變換的範疇，而由於稻田實際上並不是一個平面，因此 必定會面臨到部分無法對齊的問題。若要要求更精確的對齊，可以考慮非線性變換，

將影像切割成數塊三角形並各自變形。

參考文獻

1. Wrigley, C., Batey, I., and Miskelly, D., Chapter 25 - Grain Quality: The Future is With the Consumer, the Scientist and the Technologist, in Cereal Grains (Second Edition), C. Wrigley, I. Batey, and D. Miskelly, Editors. 2017: Woodhead Publishing. p. 691-725.

2. Leonard, E.C., Precision Agriculture, in Encyclopedia of Food Grains (Second Edition), C. Wrigley, et al., Editors. 2016, Oxford: Academic Press. p. 162-167.

3. J. P. R, B., et al. Aerial mapping of rice crops using mosaicing techniques for vegetative index monitoring. in 2018 International Conference on Unmanned Aircraft Systems (ICUAS). 2018.

4. Ahmed, I., et al. Automatic Detection and Segmentation of Lentil Crop Breeding Plots From Multi-Spectral Images Captured by UAV-Mounted Camera. in 2019 IEEE Winter Conference on Applications of Computer Vision (WACV). 2019.

5. Murata, K., et al. A Study on Growth Stage Classification of Paddy Rice by CNN using NDVI Images. in 2019 Cybersecurity and Cyberforensics Conference (CCC). 2019.

6. Lee, H., Seo, W., and Lee, K. Detection of Oak Wilt Disease Using Convolutional Neural Network From Uav Natural Color Imagery. in IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium. 2019.

7. Broto, P.E., Saputro, A.H., and Kushardono, D. Prediction of paddy field area base on aerial photography using multispectral camera. in 2017 International Conference on Sustainable Information Engineering and Technology (SIET).

2017.

8. Quiroz, R.A.A., Guidotti, F.P., and Bedoya, A.E. A method for automatic identification of crop lines in drone images from a mango tree plantation using segmentation over YCrCb color space and Hough transform. in 2019 XXII Symposium on Image, Signal Processing and Artificial Vision (STSIVA). 2019.

9. Rosa, L.E.C.L., et al. FCRN-Based Multi-Task Learning for Automatic Citrus Tree Detection From UAV Images. in 2020 IEEE Latin American GRSS & ISPRS Remote Sensing Conference (LAGIRS). 2020.

10. 葉怡萱, 基於精緻農業空拍影像全景拼接與稻田植株識別, in 工程科學及 海洋工程學研究所. 2019, 國立臺灣大學. p. 1-44.

11. Xue, J. and Su, B., Significant Remote Sensing Vegetation Indices: A Review of Developments and Applications. Journal of Sensors, 2017. 2017: p. 1353691.

12. Fuentes-Peailillo, F., et al. Comparison of vegetation indices acquired from RGB and Multispectral sensors placed on UAV. in 2018 IEEE International Conference on Automation/XXIII Congress of the Chilean Association of Automatic Control (ICA-ACCA). 2018.

13. Matsushita, B., et al., Sensitivity of the Enhanced Vegetation Index (EVI) and Normalized Difference Vegetation Index (NDVI) to Topographic Effects: A Case Study in High-density Cypress Forest. Sensors (Basel, Switzerland), 2007. 7(11):

p. 2636-2651.

14. Gitelson, A.A., et al., Novel algorithms for remote estimation of vegetation fraction. Remote Sensing of Environment, 2002. 80(1): p. 76-87.

15. Hunt, E.R., et al., A visible band index for remote sensing leaf chlorophyll content at the canopy scale. International Journal of Applied Earth Observation and Geoinformation, 2013. 21: p. 103-112.

16. Woebbecke, D.M., et al., Color indices for weed identification under various soil, residue, and lighting conditions. Transactions of the American Society of Agricultural Engineers, 1995. 38(1): p. 259-269.

17. Hamuda, E., Glavin, M., and Jones, E., A survey of image processing techniques for plant extraction and segmentation in the field. Computers and Electronics in Agriculture, 2016. 125: p. 184-199.

18. Woebbecke, D.M., et al. Plant species identification, size, and enumeration using machine vision techniques on near-binary images. in Proceedings of SPIE - The International Society for Optical Engineering. 1993.

19. Hague, T., Tillett, N.D., and Wheeler, H., Automated crop and weed monitoring in widely spaced cereals. Precision Agriculture, 2006. 7(1): p. 21-32.

20. Meyer, G.E., Hindman, T.W., and Lakshmi, K. Machine vision detection parameters for plant species identification. in Proc.SPIE. 1999.

21. Kataoka, T., et al. Crop growth estimation system using machine vision. in Proceedings 2003 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM 2003). 2003.

22. Meyer, G.E. and Neto, J.C., Verification of color vegetation indices for automated crop imaging applications. Computers and Electronics in Agriculture, 2008. 63(2):

p. 282-293.

23. Guijarro, M., et al., Automatic segmentation of relevant textures in agricultural images. Computers and Electronics in Agriculture, 2011. 75(1): p. 75-83.

24. Burgos-Artizzu, X.P., et al., Real-time image processing for crop/weed

discrimination in maize fields. Computers and Electronics in Agriculture, 2011.

75(2): p. 337-346.

25. Guerrero, J.M., et al., Support Vector Machines for crop/weeds identification in maize fields. Expert Systems with Applications, 2012. 39(12): p. 11149-11155.

26. Pizer, S.M., et al., Adaptive histogram equalization and its variations. Computer Vision, Graphics, and Image Processing, 1987. 39(3): p. 355-368.

27. Lindeberg, T., Image Matching Using Generalized Scale-Space Interest Points.

Journal of Mathematical Imaging and Vision, 2015. 52(1): p. 3-36.

28. Grady, L., Random Walks for Image Segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006. 28(11): p. 1768-1783.

29. Chebrolu, N., Läbe, T., and Stachniss, C., Robust Long-Term Registration of UAV Images of Crop Fields for Precision Agriculture. IEEE Robotics and Automation Letters, 2018. 3(4): p. 3097-3104.

30. Otsu, N., A Threshold Selection Method from Gray-Level Histograms. IEEE Transactions on Systems, Man, and Cybernetics, 1979. 9(1): p. 62-66.

31. Kotu, V. and Deshpande, B., Chapter 4 - Classification, in Data Science (Second Edition), V. Kotu and B. Deshpande, Editors. 2019: Morgan Kaufmann. p. 65-163.

32. Achanta, R., et al., SLIC Superpixels Compared to State-of-the-Art Superpixel Methods. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012.

34(11): p. 2274-2282.

33. Blob detection. Wikipedia [Online]; Available from:

https://en.wikipedia.org/w/index.php?title=Blob_detection.

34. Lowe, D.G., Distinctive Image Features from Scale-Invariant Keypoints.

International Journal of Computer Vision, 2004. 60(2): p. 91-110.

35. Bradski, G., The openCV library. Dr. Dobb's Journal of Software Tools, 2000. 25.

36. He, H.J., Zheng, C., and Sun, D.W., Chapter 2 - Image Segmentation Techniques, in Computer Vision Technology for Food Quality Evaluation (Second Edition), D.-W. Sun, Editor. 2016, San Diego: Academic Press. p. 45-63.

37. Masmoudi, L., Zennouhi, R., and El Ansari, M., Image Segmentation Based on a Two-Dimensional Histogram. 2011.

38. OpenCV: Histogram Comparison. [Online]; Available from:

https://docs.opencv.org/3.4.2/d8/dc8/tutorial_histogram_comparison.html.

39. Neubert, P. and Protzel, P. Compact Watershed and Preemptive SLIC: On Improving Trade-offs of Superpixel Segmentation Algorithms. in 2014 22nd International Conference on Pattern Recognition. 2014.

40. Gao, H., et al., Marker-based image segmentation relying on disjoint set union.

Signal Processing: Image Communication, 2006. 21(2): p. 100-112.

41. Felzenszwalb, P.F. and Huttenlocher, D.P., Efficient Graph-Based Image Segmentation. International Journal of Computer Vision, 2004. 59(2): p. 167-181.

42. McCool, M., Robison, A.D., and Reinders, J., Chapter 11 - K-Means Clustering, in Structured Parallel Programming, M. McCool, A.D. Robison, and J. Reinders, Editors. 2012, Boston: Morgan Kaufmann. p. 279-289.

43. Tareen, S.A.K. and Saleem, Z. A comparative analysis of SIFT, SURF, KAZE, AKAZE, ORB, and BRISK. in 2018 International Conference on Computing, Mathematics and Engineering Technologies (iCoMET). 2018.

44. Bay, H., et al., Speeded-Up Robust Features (SURF). Computer Vision and Image Understanding, 2008. 110(3): p. 346-359.

45. Alcantarilla, P.F., Bartoli, A., and Davison, A.J. KAZE Features. in Computer Vision – ECCV 2012. 2012. Berlin, Heidelberg: Springer Berlin Heidelberg.

46. Fernández Alcantarilla, P., Fast Explicit Diffusion for Accelerated Features in Nonlinear Scale Spaces. 2013.

47. Rublee, E., et al. ORB: An efficient alternative to SIFT or SURF. in 2011 International Conference on Computer Vision. 2011.

48. Leutenegger, S., Chli, M., and Siegwart, R.Y. BRISK: Binary Robust invariant scalable keypoints. in 2011 International Conference on Computer Vision. 2011.

49. Random sample consensus, in Wikipedia. 2020.

50. van der Walt, S., et al., scikit-image: Image processing in Python. PeerJ, 2014. 2.

51. Tremeau, A. and Colantoni, P., Regions adjacency graph applied to color image segmentation. IEEE Transactions on Image Processing, 2000. 9(4): p. 735-744.

52. Wilcox, R.R., 6 - LEAST SQUARES REGRESSION AND PEARSON'S CORRELATION, in Applying Contemporary Statistical Techniques, R.R. Wilcox, Editor. 2003, Burlington: Academic Press. p. 173-206.

53. Hartley, R. and Zisserman, A., Multiple View Geometry in Computer Vision. 2003:

Cambridge University Press.

54. Nixon, M.S. and Aguado, A.S., Chapter 10 - Appendix 1: Camera geometry fundamentals, in Feature Extraction & Image Processing for Computer Vision (Third Edition), M.S. Nixon and A.S. Aguado, Editors. 2012, Oxford: Academic Press. p. 489-518.

55. Medhi, D. and Ramasamy, K., Chapter 2 - Routing Algorithms: Shortest Path, Widest Path, and Spanning Tree, in Network Routing (Second Edition), D. Medhi

and K. Ramasamy, Editors. 2018, Boston: Morgan Kaufmann. p. 30-63.

56. 裴昭如, 透過無人機早期多光譜影像判斷水稻植株位置, in 臺灣大學農藝 學研究所學位論文. 2019, 國立臺灣大學. p. 1-60.

附錄

表 17 Fz 之稻田區域影像分割結果

影像編號 分割稻田面積 (像素)

標記稻田面積 (像素)

TP 面積

(像素) precision recall f-measure DSC07281 1.87E+07 1.89E+07 1.86E+07 99.34% 98.17% 98.76%

DSC07283 1.53E+07 1.57E+07 1.51E+07 98.67% 96.20% 97.42%

DSC07285 1.14E+07 1.13E+07 1.12E+07 97.89% 98.62% 98.25%

DSC07293 8.04E+06 8.07E+06 7.75E+06 96.39% 96.07% 96.23%

DSC07299 1.72E+07 1.74E+07 1.70E+07 98.73% 97.72% 98.22%

DSC07300 1.73E+07 1.66E+07 1.64E+07 94.94% 98.72% 96.80%

DSC07323 1.79E+07 1.72E+07 1.69E+07 94.37% 98.43% 96.36%

DSC07325 1.91E+07 1.77E+07 1.75E+07 92.11% 99.10% 95.48%

DSC07327 1.60E+07 1.54E+07 1.50E+07 93.93% 97.43% 95.65%

DSC07328 1.00E+07 9.78E+06 9.62E+06 96.18% 98.37% 97.26%

表 18 SL_R 之稻田區域影像分割結果

表 20 CLAHE 之秧苗偵測結果

稻田區域 偵測數量 實際數量 TP 數量 precision recall f-measure 3 228 244 193 84.65% 79.10% 81.78%

4 251 269 204 81.27% 75.84% 78.46%

7 273 249 196 71.79% 78.71% 75.10%

8 283 244 222 78.45% 90.98% 84.25%

10 207 226 188 90.82% 83.19% 86.84%

12 236 226 208 88.14% 92.04% 90.04%

表 21 ExG 之秧苗偵測結果

稻田區域 偵測數量 實際數量 TP 數量 precision recall f-measure 3 323 244 237 73.37% 97.13% 83.60%

4 305 269 268 87.87% 99.63% 93.38%

7 378 249 243 64.29% 97.59% 77.51%

8 243 244 240 98.77% 98.36% 98.56%

10 275 226 217 78.91% 96.02% 86.63%

12 291 226 221 75.95% 97.79% 85.49%

表 22 ExR 之秧苗偵測結果

稻田區域 偵測數量 實際數量 TP 數量 precision recall f-measure 3 254 244 228 89.76% 93.44% 91.57%

4 300 269 260 86.67% 96.65% 91.39%

7 256 249 221 86.33% 88.76% 87.52%

8 278 244 241 86.69% 98.77% 92.34%

10 211 226 204 96.68% 90.27% 93.36%

12 189 226 182 96.30% 80.53% 87.71%

表 23 MExG 之秧苗偵測結果

稻田區域 偵測數量 實際數量 TP 數量 precision recall f-measure 3 259 244 231 89.19% 94.67% 91.85%

4 256 269 249 97.27% 92.57% 94.86%

7 202 249 194 96.04% 77.91% 86.03%

8 247 244 231 93.52% 94.67% 94.09%

10 192 226 190 98.96% 84.07% 90.91%

12 188 226 185 98.40% 81.86% 89.37%

表 24 TGI 之秧苗偵測結果

稻田區域 偵測數量 實際數量 TP 數量 precision recall f-measure 3 250 244 234 93.60% 95.90% 94.74%

4 273 269 266 97.44% 98.88% 98.15%

7 201 249 189 94.03% 75.90% 84.00%

8 252 244 241 95.63% 98.77% 97.18%

10 214 226 209 97.66% 92.48% 95.00%

12 206 226 203 98.54% 89.82% 93.98%

表 25 COM2 之秧苗偵測結果

稻田區域 偵測數量 實際數量 TP 數量 precision recall f-measure 3 250 244 229 91.60% 93.85% 92.71%

4 286 269 265 92.66% 98.51% 95.50%

7 192 249 188 97.92% 75.50% 85.26%

8 252 244 239 94.84% 97.95% 96.37%

10 203 226 201 99.01% 88.94% 93.71%

12 196 226 193 98.47% 85.40% 91.47%