2024 Fiscal Year Final Research Report
Development of a method for separating pattern noise in images and characterizing pattern components using frequency spectrum
| Project/Area Number |
22K04123
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 21030:Measurement engineering-related
|
|---|
| Research Institution | Shinshu University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2022-04-01 – 2025-03-31
|
| Keywords | 自己相関 / ピーク解析 / パタン解析 / スペクトル解析 |
|---|
| Outline of Final Research Achievements |
This study aims to develop a method for removing a pattern component when an unnecessary repetitive pattern image appears in the desired image. To improve its performance, we developed a method to estimate the “pattern size”, and incorporated the method into our pattern removal method, and presented the results at domestic research workshops.
In the pattern removal method, the spectrum in the Fourier transformed domain is used to characterize the pattern component. Estimating the pattern size allows for an appropriate selection of the data length for Fourier transformation, which enhances the spectral characteristics and facilitates the separation of pattern components. In the previous version of the pattern removal method, pattern components frequently remained around the image edges. On the other hand, after the integration of the proposed method, these pattern components were substantially reduced.
|
| Free Research Field |
画像処理
|
| Academic Significance and Societal Importance of the Research Achievements |
パタンノイズの除去にも利用できるが，信号処理やパタン認識などの分野で何らか信号列の周期を推定しようとする試みにも利用できる可能性がある．今回の手法では，自己相関の利用による「ピーク値の間隔を推定する問題」への変換，及び，「ピーク値に接する接線の傾きを推定する問題」への変換という２つの提案を行った．「ピーク値の間隔を推定する」という問題は，信号処理やパタン認識の分野でも度々扱われるが，ピーク位置の検出，及び，クラスタリングが関係し，人間には簡単に知覚できるが，計算によって算出することが難しい問題である．今回の数値計算においては，この難しさを念頭において，問題を簡単化できるように設計している．
|
