Toning up Images by Smoothening Edges

dc.contributor.authorGatoto, James
dc.contributor.authorFwamba, Rukia Nasimiyu
dc.date.accessioned2025-10-27T09:54:53Z
dc.date.available2025-10-27T09:54:53Z
dc.date.issued2024
dc.description.abstractBackground: Under Partial Differential Equations, an image is a function 2 f x y X ( , ),  . These equations are used in smoothening of images. The necessity to have an excellent image quality is increasingly required in the current world. Most of the obtained images are not as smooth as we would want hence they are blurred. Use of the nonlinear diffusion equation is essential in the current day smoothening of images. This model inputs smoothness in the denoising process. This research improved the quality of images through the use of nonlinear PDEs of parabolic nature i.e. the heat equation. Materials and Methods: Numerical schemes of ADI (Alternating Direction Implicit method) and 2-EGSOR(2- Point Explicit Group Successive Over Relaxation) were used to solve the equations in MATLAB subject to an initial condition of a noisy image, generating various smoothened images. Results: Comparatively output from ADI is very close to the original image in terms of alignment, smoothness i.e. refined texture, well outlined contours and overall detail without stair casing. .This method is characterized with a smaller residue and much time lapse. An error analysis too was carried out using Root Mean Square Method. ADI & the blocked (ADI and EGSOR) register a comparatively lower RMSE. Conclusion: The most suitable algorithm for image smoothening is Alternating Direction Implicit Method
dc.identifier.citationFwamba, R. N., & Gatoto, J. (2024). Toning up images by smoothening edges. IOSR Journal of Mathematics (IOSR-JM), 20(1, Ser. 4), 14–24. https://doi.org/10.9790/0661-2001041424
dc.identifier.issn2278-8727
dc.identifier.urihttps://ir-library.ku.ac.ke/handle/123456789/31851
dc.language.isoen
dc.publisherIOSR-JM
dc.titleToning up Images by Smoothening Edges
dc.typeArticle
Files
Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
Full-text Research Article.pdf
Size:
1.21 MB
Format:
Adobe Portable Document Format
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
2.66 KB
Format:
Item-specific license agreed upon to submission
Description: