A generalised framework for saliency-based point feature detection
- Submitting institution
-
Middlesex University
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 773
- Type
- D - Journal article
- DOI
-
10.1016/j.cviu.2016.09.008
- Title of journal
- Computer Vision and Image Understanding
- Article number
- -
- First page
- 117
- Volume
- 157
- Issue
- -
- ISSN
- 1077-3142
- Open access status
- Compliant
- Month of publication
- September
- Year of publication
- 2016
- URL
-
http://eprints.mdx.ac.uk/20575/
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
2
- Research group(s)
-
-
- Citation count
- 4
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Computer vision systems, are frequently required to identify matching features in a manner that is robust to changes in view-point, lighting, etc. Certain situations also require this matching to take place across different instrumental modalities, for example, pairing image-features with 3D data obtained by LiDAR scanners. In this case, features must generalise across modalities. This paper's significance lies in its proposal of a completely novel information-theory based method for determining universal, modality-independent features. Experimental results conducted over a large variety of datasets demonstrate that the method significantly outperforms the state-of-the-art, with potentially universal application to arbitrary imaging domains.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -