Spline-based Approach for Image Restoration

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：18

、訪客IP：18.191.27.94

姓名

何昭明(Zhao-Ming He) 查詢紙本館藏

畢業系所

統計研究所

論文名稱

(Spline-based Approach for Image Restoration)

相關論文

★ A Compression-Based Partitioning Estimate Classifier	★ Data adaptive median filters for image denoising based on a prediction criterion
★ Fixed effect estimation and spatial prediction via universal kriging	★ Two-stage model selection under a misspecified spatial covariance function
★ 時空過程的配適研究	★ 空間變異係數模型
★ 非監督式廣義學習NEM分類演算法	★ 固定秩克里金法的圖像重建
★ 高維資料空間零膨脹模型的有效參數估計	★ 克利金模型中基於Kullback-Leibler損失的共變異函數選擇
★ A Spatio-temporal Hierarchical PGEV Model for Extreme Value Analysis

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 (2026-7-1以後開放)

摘要(中)

在影像研究中，影像的底層結構通常呈現出空間非平穩特徵。適當地指定空間相關性的非平穩函數以重建影像是一個具有挑戰性的問題。此外，重建高解析度影像的計算問題也是另一項挑戰，因為這可能耗時甚至無法執行。在本篇論文中，我們運用薄板樣條（Thin-plate spline）技術在空間迴歸模型中對影像的空間相關性進行建模。我們提出均方預測誤差（MSPE）準則，評估並選取一個適當的抽樣方法和適中的樣本數，為了充分降維並重建感興趣的影像。我們透過各種模擬情境來評估所提出的MSPE準則的有效性，並以氣象和醫學影像的兩個實際數據說明方法的實用性。

摘要(英)

In image studies, the underlying structure of images generally shows spatially nonstationary features. Appropriately specifying a nonstationary covariance function for the inherent spatial correlation to reconstruct the underlying image of interest is a challenging problem. In addition, the computational issue about reconstructing high-resolution images is also another challenge, as it can be time-consuming or even infeasible. In this thesis, we apply the thin-plate spline technique to model the underlying spatial correlation of images within spatial regression models. To achieve dimension reduction, we then propose a mean squared prediction error (MSPE) criterion to determine an appropriate sampling method with a moderate sample size to reconstruct the underlying image of interest. We assess the empirical performance of the proposed MSPE criterion via various simulation scenarios and two real data examples regarding meteorology and medical imaging are presented for illustration.

關鍵字(中)

★ 資料擾動
★ 廣義自由度
★ 分割
★ 超級壓縮
★ 平滑樣條

關鍵字(英)

★ Data perturbation
★ Generalized degrees of freedom
★ SPlit
★ Supercompress
★ Thin-plate spline

論文目次

1 Introduction 1
2 Spatial regression model for images 4
2.1 Spatial regression model 4
2.2 Thin-plate splines 5
2.3 Parameter estimation 6
2.4 Image restoration 8
3 Sampling methods 9
3.1 SPlit 9
3.2 Supercompress 12
4 Determination of sample size and sampling method 16
4.1 Generalized degrees of freedom 16
4.2 Data perturbation 17
4.3 Selection criterion and optimality 18
5 Simulation study 20
5.1 Settings 20
5.2 Results 23
6 Real data study 28
6.1 Estimation of nugget effect 28
6.2 Meteorology example 29
6.3 CT image example 35
7 Conclusion and discussion 38
Reference 40

參考文獻

1. Cressie, N. (1993). Statistics for Spatial Data (revised edition), Wiley: New York.

2. Cressie, N. and Hawkins, D.M. (1980). Robust estimation of the variogram. Journal of the
International Association for Mathematical Geology, 12(2), 115-125.

3. Huang, H. C., and Chen, C. S. (2007). Optimal Geostatistical Model Selection. Journal of
the American Statistical Association, 102, 1009-1024.

4. Joseph, V. R., and Mak, S. (2021). Supervised compression of big data. Stat Anal Data
Min: The ASA Data Sci Journal, 14, 217-229.

5. Joseph, V. R., and Vakayil, A. (2021). SPlit: An Optimal Method for Data Splitting.
Taylor and Francis Group, LLC, 0, 1-11.

6. Lo`eve, M. (1978). Probability Theory (fourth edition), Springer: New York.

7. Ludwig, G., Zhu, J., Reyes, P., Chen, C. S., and Conley, S. P. (2020). On spline-based
approaches to spatial linear regression for geostatistical data. Environmental and Eco-
logical Statistics, 27, 175–202.

8. MacQueen, J. (1967). Some Methods for Classification and Analysis of Multivariate Obser-
vations. Proceedings of the 5th Berkeley Symposium on Mathematical Statistics and
Probability University of California Press, 1, 281-297.

9. Mat ́ern, B. (1986). Spatial Variation (second edition), Springer: New York

10. Ramsay, J.O., and Silverman, B.W. (2005). Functional Data Analysis, Springer: New York.

11. Stroup, W. W. (2012). Generalized Linear Mixed Models: Modern Concepts, Methods and
Applications, CRC Press.

12. Szekely, G.J., and Rizzo, M.L. (2013). Energy Statistics: A Class of Statistics Based on
Distances. Journal of Statistical Planning and Inference, 143, 1249-1272.

13. Wahba, G. (1990). Spline Models for Observational Data, Philadelphia: Society for Indus-
trial and Applied Mathematics.

14. Wahba, G., and Wold, S. (1975). A completely automatic french curve: fitting spline
functions by cross validation. Communications in Statistics, 4, 1-17.

指導教授

陳春樹(Chun-Shu Chen)

審核日期

2023-7-14

推文