LANDSAT 위성영상을 이용한 벼 생육 및 수량 모니터링 (A Study on Rice Growth and Yield Monitoring Using Medium Resolution Landsat Imagery)

서지정보

· 발행기관 : 한국작물학회
· 수록지 정보 : 한국작물학회지 / 53권 / 4호 / 388 ~ 393페이지
· 저자명 : 김민호, 이충근, 박호기, 이재은, 구본철, 신진철

초록

Earth observation satellite imagery having
medium-resolution can provide the useful information very
rapidly and cheaply. The objective of this study was to
assess the feasibility for monitoring rice growth and yield
using medium resolution satellite imagery at Seosan AB
reclaimed area, Chung-nam province.
Using the LANDSAT imagery at booting stage (29th July
2004), NDVIR had the most significant linear relationships
with rice yield of Seosan AB reclaimed area with the correlation
coefficient (r) as 0.68. Therefore, this relationship
was established as rice yield equation as function of
NDVIR, where excluding the 10 small area having low
number of pixel, the determination coefficient (R²) of the
linear regression between NDVIred and milled rice yield
was improved to 0.66. In addition, raster masking method,
which was easier and faster even if a little unaccurate than
preexisting method, was established for extracting information
paddy field zone. Adaptability of rice yield equation
function of NDVIR on year and region was investigated
using rice yield and NDVIR values, which were extracted
with raster masking method, from 7 counties or cities,
Kyeong-ki province in 2005. Relationship between observed
and calculated rice yield showed 1:1 line indicating that the
adaptability was admitted.

영어초록

Earth observation satellite imagery having
medium-resolution can provide the useful information very
rapidly and cheaply. The objective of this study was to
assess the feasibility for monitoring rice growth and yield
using medium resolution satellite imagery at Seosan AB
reclaimed area, Chung-nam province.
Using the LANDSAT imagery at booting stage (29th July
2004), NDVIR had the most significant linear relationships
with rice yield of Seosan AB reclaimed area with the correlation
coefficient (r) as 0.68. Therefore, this relationship
was established as rice yield equation as function of
NDVIR, where excluding the 10 small area having low
number of pixel, the determination coefficient (R²) of the
linear regression between NDVIred and milled rice yield
was improved to 0.66. In addition, raster masking method,
which was easier and faster even if a little unaccurate than
preexisting method, was established for extracting information
paddy field zone. Adaptability of rice yield equation
function of NDVIR on year and region was investigated
using rice yield and NDVIR values, which were extracted
with raster masking method, from 7 counties or cities,
Kyeong-ki province in 2005. Relationship between observed
and calculated rice yield showed 1:1 line indicating that the
adaptability was admitted.

참고자료

· 없음