On ear-based human identification in the mid-wave infrared spectrum
Image and Vision Computing
| Hi-index | 0.00 |
We present an experimental study to demonstrate the effect of the time difference in image acquisition for gallery and probe on the performance of ear recognition. This experimental research is the first study on the time effect on ear biometrics. For the purpose of recognition, we convolve banana wavelets with an ear image and then apply local binary pattern on the convolved image. The histograms of the produced image are then used as features to describe an ear. A histogram intersection technique is then applied on the histograms of two ears to measure the ear similarity for the recognition purposes. We also use analysis of variance (ANOVA) to select features to identify the best banana wavelets for the recognition process. The experimental results show that the recognition rate is only slightly reduced by time. The average recognition rate of 98.5% is achieved for an eleven month-difference between gallery and probe on an un-occluded ear dataset of 1491 images of ears selected from Southampton University ear database.