상세 보기
CMR-OV 초분광 수심 기법의 최대측정가능수심 확장을 통한 국내 대하천 정밀하상측정 가능성 평가
- 김도현;
- 김동수;
- 유호준;
- 이서경;
- 권시윤
SCOPUS
0초록
River bathymetry varies spatially and temporally in response to discharge fluctuations and bed movement, making accurate depth estimation important for understanding and managing river systems. However, conventional field surveys and contact-based measurements are limited to point-based observations, making it difficult to obtain continuous bathymetric information over large river reaches. To address these limitations, this study applied the machine learning–based CMR-OV algorithm to hyperspectral image–based bathymetry and evaluated its potential to extend the maximum measurable depth. Depth data measured using an ADCP were grouped at 0.5 m intervals for model training, and the applicable depth range was evaluated based on the model showing the lowest RMSE. In addition, a bathymetric map was generated from hyperspectral imagery using the trained model, and its performance was evaluated by comparing predicted depths with observed depths along validation cross-sections. The results showed that, under low-turbidity conditions, CMR-OV achieved relatively good performance over a deeper depth range than the 2.5 m maximum measurable depth previously reported for the conventional OBRA approach. Cross-sectional comparison showed the best performance with an RMSE of 0.21 m and a coefficient of determination (R2) of 0.96. However, because in-situ observations in deeper water were limited, it was difficult to define the maximum measurable depth as a single fixed value. Further investigation using deeper-water observations and independent validation datasets is therefore needed.
키워드
- 제목
- CMR-OV 초분광 수심 기법의 최대측정가능수심 확장을 통한 국내 대하천 정밀하상측정 가능성 평가
- 제목 (타언어)
- Bed morphology measurement capability in large Korean rivers extending maximum measurable depth using CMR-OV hyperspectral bathymetry method
- 저자
- 김도현; 김동수; 유호준; 이서경; 권시윤
- 발행일
- 2026-08
- 유형
- Y
- 저널명
- 한국수자원학회 논문집
- 권
- 59
- 호
- 8
- 페이지
- 907 ~ 920
- 언어
- KOR
- 출판사
- 한국수자원학회
- 발행국가
- 대한민국
- 분량
- 14 페이지
- ISSN
- E 2799-8754
P 2799-8746