표제지
목차
Abstract 10
제1장 서론 12
1.1. 연구배경 및 목적 12
1.2. 연구내용 및 방법 14
제2장 해안선 변화 모니터링을 위한 위성정보 활용 및 객체추출 방법 16
2.1. 해안선 변화 모니터링 관측을 위한 국내외 위성정보 활용 현황 16
2.2. 고해상도 영상 기반 해안선 매핑 기법 개발에 관한 국내외 연구 동향 18
제3장 다중시기 아리랑 위성영상을 활용한 해안선 변화 모니터링 시스템 구축 26
3.1. 위성영상 기반 해안선 추출 알고리즘 구축을 위한 적용 이론 검토 26
3.1.1. 정규수분지수(NDVI: Normalized Difference Water Index) 27
3.1.2. 정규식생지수(NDVI: Normalized Difference Vegetation Index) 29
3.1.3. 에지 검출 기법(Edge Detection Technique) 31
3.1.4. 이진화 기법(Thresholding Technique) 34
3.1.5. 머신러닝(기계학습) 기법(Machine Learning Technique) 38
3.1.6. 모폴로지 필터링(Morphological Fitlering) 42
3.2. 위성영상을 활용한 해안선 추출 및 변화 모니터링 시스템 구축 45
3.2.1. 위성영상을 활용한 해안선 추출 알고리즘 설계 45
3.2.2. 해안선 자동 추출 알고리즘 개발 48
3.2.3. 해안선 매핑 프로토타입 개발 52
3.2.4. 해안선 매핑 및 변화 모니터링 시스템 개발 57
제4장 해안선 추출 및 변화 모니터링 시스템을 활용한 해안선 변화 분석 및 정확도 검증 67
4.1. 정확도 검증을 위한 적용 지역 선정 67
4.2. 추출결과의 정확도 검증 68
제5장 결론 71
References 75
Table 1. Comparison of shoreline extraction result accuracy 69
Fig. 1. Plural segments (left) created by applying mean-shiht segmentation... 18
Fig. 2. Manually modified/edited outer boundary of the segment,... 19
Fig. 3. Outcome (right) of the application of the aerial photo (left)... 20
Fig. 4. Edges extracted using the edge extraction technique 20
Fig. 5. Process of determining the shoreline height based on LiDAR DSM... 21
Fig. 6. Test-bed coastal area of Erie Lake in the U.S. in this study 22
Fig. 7. 3D shoreline based on satellite images and aerial LiDAR data obtained... 23
Fig. 8. 3D shoreline based on satellite images and aerial LiDAR data obtained... 24
Fig. 9. NDWI images produced using the Green/NIR band 28
Fig. 10. NDVI images (right) produced using the multiple spectral bands... 30
Fig. 11. Edge group (right) extracted from a high-resolution image (left)... 33
Fig. 12. Shoreline data extracted from NDWI images using the binarization technique 36
Fig. 13. Analysis of land use situation by applying the machine learning... 41
Fig. 14. Process of extracting precise shoreline data by applying morphology... 44
Fig. 15. Shoreline data extraction algorithm design plan... 45
Fig. 16. System concept for shoreline automatic mapping 46
Fig. 17. Clustering outline for Vectorizing (CV) algorithm 49
Fig. 18. Shoreline edge detection (SED) algorithm outline 50
Fig. 19. Outline of development of algorithms for extracting shoreline data... 51
Fig. 20. Execution screen of the proposed shoreline mapping prototype 52
Fig. 21. Process of inputting the Arirang satellite image band in the... 53
Fig. 22. Arirang satellite images inputted in the shoreline mapping prototype 53
Fig. 23. Binarized image produced from the shoreline mapping prototype... 54
Fig. 24. Borderline detected in the shoreline mapping prototype based on... 54
Fig. 25. Raster-format shoreline data extracted... 55
Fig. 26. Vector-format shoreline data finally extracted using the shoreline... 56
Fig. 27. Outline of the shoreline automatic mapping... 57
Fig. 28. Process of loading Arirang satellite images from the Korean... 58
Fig. 29. Mini-map activated from the integrated system 58
Fig. 30. Precess of expanding and reducing the test-bed area through the... 59
Fig. 31. Process of applying the ANN technique to Arirang satellite images... 60
Fig. 32. Water-land cluster created by applying the ANN technique to... 60
Fig. 33. Process of producing vector-format shoreline data using the... 61
Fig. 34. Process of loading the produced vector-format shoreline shape files... 61
Fig. 35. Outcome of the vector-format shoreline shape files loaded onto the... 62
Fig. 36. Vector-format shoreline shape file obtained using QGIS software 62
Fig. 37. Two shoreline data extracted from multiple-time Arirang satellite... 63
Fig. 38. SHP comparison window loaded from the integrated system to... 64
Fig. 39. Process of inputting the two shoreline data extracted from... 64
Fig. 40. Three screens created by inputting the two shoreline data... 65
Fig. 41. Outcome of the average travel distance of the two shoreline data... 65
Fig. 42. Result of the application of the shoreline extraction algorithm to... 67
Fig. 43. Process of verifying the accuracy of the shoreline extraction output 68
Fig. 44. Comparison of ANN-based shoreline extraction accuracy 69