2022 Fiscal Year Final Research Report
Analysis of subsurface structure of biogical tissue using spatial light transport
| Project/Area Number |
20K19825
|
|---|
| Research Category |
Grant-in-Aid for Early-Career Scientists
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
Basic Section 61010:Perceptual information processing-related
|
|---|
| Research Institution | Kyushu University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2020-04-01 – 2023-03-31
|
| Keywords | コンピュータビジョン / 光伝播解析 / 深層学習 / ドメイン適応 |
|---|
| Outline of Final Research Achievements |
The purpose of this study is to visualize and reconstruct the internal structure of biological tissue, which has been difficult to analyze due to light scattering, using optical measurement methods based on computer vision and machine learning. Through this study, we proposed a method for measuring the internal structure of scatterers by using patterned light projection using a projector, and an algorithm for estimating the structure from observed images. In addition, we proposed a method for generating an efficient dataset for machine learning, which can learn models that can withstand real-world measurements while using synthetic data.
|
| Free Research Field |
コンピュータビジョン
|
| Academic Significance and Societal Importance of the Research Achievements |
光散乱は本研究で主として取り上げた生体のほか，プラスチックや植物など多くの半透明物体で起こる現象である．提案する光学計測手法はカメラやプロジェクタで実現可能であるため，手軽に計測環境を構築可能であり，家庭内における簡易な診断や生産現場での検査など幅広い応用が期待できる．また，物理モデルを考慮した合成画像データセット構築手法は，合成データによる学習で実環境利用を実現するものであり，深層機械学習の応用範囲の拡大に寄与するものである．
|
