Abstract | ||
---|---|---|
Principal Component Analysis (PCA) has been successfully used for many application including ear recognition. However, its performance is limited due to its significant data dependency. This paper presents a two dimensional multi-band PCA (2D-MBPCA) method, which has shown a significantly higher performance to that of the PCA. The proposed method divided the input gray image into a number of images, based on the intensity of its pixels using either a dynamic or predefined equal range of threshold values. PCA is then applied on the resulting set of images to extract their features. The resulting features are used to find the best match. The application of the proposed 2D-MBPCA for ear recognition using two benchmark ear image datasets, shows the merit of the proposed technique to that of the standard PCA. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1109/IST.2018.8577132 | 2018 IEEE International Conference on Imaging Systems and Techniques (IST) |
Keywords | Field | DocType |
PCA,ear recognition,histogram equalization | Ear recognition,Computer vision,Histogram,Data dependency,Pattern recognition,Multi band,Computer science,Hyperspectral imaging,Feature extraction,Pixel,Artificial intelligence,Principal component analysis | Conference |
ISSN | ISBN | Citations |
1558-2809 | 978-1-5386-6629-6 | 0 |
PageRank | References | Authors |
0.34 | 0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Matthew Zarachoff | 1 | 0 | 1.35 |
Akbar Sheikh Akbari | 2 | 0 | 1.69 |
Dorothy Ndedi Monekosso | 3 | 61 | 11.89 |