Content area

Abstract

Image preprocessing is crucial in deep learning models for high performance and accuracy. This study implements AdaptBlur, a dynamic linear filter developed from a second-order partial differential equation using finite difference approximations, which enhances image quality while preserving the image structure with dynamic parameters that are optimized through the Nelder-Mead optimization, minimizing mean square error to improve the effectiveness of the filter. Moreover, the study evaluates the impact of the results on model performance on deep learning classification using publicly available datasets. Experimentally, image classifiers trained on preprocessed data with the AdaptBlur filter perform much better than those trained without filtering or with filtering using the conventional Gaussian filter, as this filter gives nearly a 10% increase in classification accuracy.

Details

1010268
Title
AdaptBlur: Adaptive Linear Filter for Enhanced Deep Learning Classification Performance
Number of pages
97
Publication year
2025
Degree date
2025
School code
0156
Source
MAI 86/12(E), Masters Abstracts International
ISBN
9798280720350
Committee member
Bevelacqua, Anthony; Prescott, Timothy
University/institution
The University of North Dakota
Department
Mathematics
University location
United States -- North Dakota
Degree
M.S.
Source type
Dissertation or Thesis
Language
English
Document type
Dissertation/Thesis
Dissertation/thesis number
32000420
ProQuest document ID
3216752328
Document URL
https://www.proquest.com/dissertations-theses/adaptblur-adaptive-linear-filter-enhanced-deep/docview/3216752328/se-2?accountid=208611
Copyright
Database copyright ProQuest LLC; ProQuest does not claim copyright in the individual underlying works.
Database
ProQuest One Academic