2022 Fiscal Year Final Research Report
Integration of graph signal processing and hyperspectral image processing
| Project/Area Number |
21K21312
|
|---|
| Research Category |
Grant-in-Aid for Research Activity Start-up
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
1002:Human informatics, applied informatics and related fields
|
|---|
| Research Institution | Tokyo Institute of Technology |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2021-08-30 – 2023-03-31
|
| Keywords | ハイパースペクトル画像処理 |
|---|
| Outline of Final Research Achievements |
To adopt graph signal processing into hyperspectral (HS) imaging, we deeply analyzed the prior knowledge on HS images. Thanks to effective evaluation of spatial architecture, our proposed method can estimate an HS image with high spatial and spectral resolution. From the results, we can see the spatial architecture helps to improve the performance for HS image restoration, so we tried to construct the methodology of spatial architecture into the graph representation for HS image restoration.
Then, we proposed a new hyperspectral imaging framework near by real imaging systems. As a result, we showed the problem can be resolved by the suitable evaluation of apriori knowledge on HS images.
|
| Free Research Field |
画像処理
|
| Academic Significance and Societal Importance of the Research Achievements |
ハイパースペクトル画像（HSI)が持つ物体固有の情報は，地球表面での異常検出に役立っており，医療や農業などのあらゆる分野が抱える課題解決に対しても重要な役割を果たすと注目されている．しかし，HSIは空間解像度が低いという問題がある．本課題では，撮影対象の空間的構造を適切に評価することで，観測の際に犠牲になっていた空間解像度を効果的に推定しており，波長情報の歪みの少ない高精細なハイパースペクトル画像の再構成を可能とした．これにより，今まで空間解像度の低さから適用を見送っていた諸分野に対してHSI利用が可能になり，様々な技術の発展に貢献すると期待される．
|
