Forest environment monitoring using texture analysis of high-resolution satellite data
| Project/Area Number |
13660142
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
林学
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
TSUYUKI Satoshi The University of Tokyo, Graduate School of Agricultural and Life Sciences, Associate Professor, 大学院・農学生命科学研究科, 助教授 (90217381)
|
|---|
| Project Period (FY) |
2001 – 2002
|
| Project Status |
Completed (Fiscal Year 2002)
|
| Budget Amount *help |
¥3,500,000 (Direct Cost: ¥3,500,000)
Fiscal Year 2002: ¥1,700,000 (Direct Cost: ¥1,700,000)
Fiscal Year 2001: ¥1,800,000 (Direct Cost: ¥1,800,000)
|
|---|
| Keywords | QuickBird / IKONOS / natural forest / segmentation / tree canopy density / mixed forest / 人工林 / 立木本数 / 樹種区分 / 林相区分 |
|---|
| Research Abstract |
It is considered very difficult to identify precise forest type classification of natural forest due to limitation of manual interpretation of aerial photograph or limitation of ground resolution of satellite imagery, though high-resolution satellite data will be the break through of such limitations. Permanent plots in the Tokyo University Forest in Hokkaido was employed as the test site of this study. IKONOS data acquired on August 16, 2001 and QuickBird data acquired on June 7, 2002 was used.
In the first year, IKONOS data was used for analysis. The statistical value of digital number of all spectral bands within each tree crown was calculated to identify tree species. Tree species were classified into two groups of broad leave trees and needle leaf trees. According to statistical test, it is well separated between groups, but it was difficult with in each group.
In the second year, QuickBird data was analyzed using textural segmentation method. It was found out that by segmentation method, individual tree crown was not recognized but several grouping level such as by log price or by tree family group could be identified.
From this study, it was clarified that large scale mixed natural forest condition can be monitored using high-resolution remote sensing data.
|
Report
(3 results)
Research Products
(8 results)
