Lorentz-modulated multiscale nonlinear diffusion for stitching in near-uniform scenes.
Saved in:
| Title: | Lorentz-modulated multiscale nonlinear diffusion for stitching in near-uniform scenes. |
|---|---|
| Authors: | Jong, Tze Kian1,2 (AUTHOR) johntkian@gmail.com, Bong, David B. L.1 (AUTHOR) davidblbong@yahoo.com |
| Source: | Multimedia Tools & Applications. May2025, Vol. 84 Issue 17, p18199-18222. 24p. |
| Subjects: | Partial differential equations, Time dilation, Artificial intelligence, Image processing, Detectors, Special relativity (Physics) |
| Abstract: | Image stitching finds diverse applications in multimedia contexts, and creating panoramic images with this technique can be particularly challenging in near-uniform scenes. Traditional feature detectors often struggle to identify distinctive features concealed within these scenes. The problem arises from the presence of featureless or homogeneous content lacking the necessary distinctiveness required to provide abundant and widely dispersed corresponding interest points. Such problems can result in unsatisfactory visual outcomes during the stitching process, manifesting as conspicuous artifacts like seams, ghosting, and geometric distortion due to insufficient matchable inliers between overlapping images. This paper presents a novel approach to feature detection, employing a nonlinear diffusion method that involves modifying the conductivity function of the partial differential equation. Inspired by the time dilation phenomenon in Einstein's theory of special relativity, we incorporate the Lorentz factor into the conductivity function, enabling the construction of novel multiscale nonlinear scale spaces that can effectively detect features in homogeneous regions and accurately stitch multiple images. Our experimental findings reveal that the proposed method consistently surpasses other state-of-the-art techniques in detecting extensive features and enhancing image stitching quality in near-uniform scenes. [ABSTRACT FROM AUTHOR] |
| Copyright of Multimedia Tools & Applications is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
Be the first to leave a comment!
