Palm Vein Recognition Using Scale Invariant Feature Transform with RANSAC Mismatching Removal

The rapid growth of technology has increased the demand for automated security systems. Due to the accessibility of the palm region and the unique characteristics of each individual's palm vein features, such biometrics have been receiving particular attention. In the published research relating to palm vein biometrics, usually only a single image is used to supply the data for recognition purposes. Previous experimental work has demonstrated that the fusion of multiple images is able to provide richer feature information resulting in an improved classification performance. However, although most of the image fusion techniques are able to preserve the vein pattern, the fused image is often blurred, the colors are distorted and the spatial resolution reduced. In this paper, the multiresolution discrete cosine transform (MRDCT) and frequency partition DCT (FPDCT) image fusion are applied and are able to extract the finer details of vein patterns while reducing the presence of noise in the image. The performance shows that the use of MRDCT and FPDCT was able to improve recognition rate compared to using a single image. The equal error rate improvement is also significant, falling to 9% in 700nm image, 7% in 850nm image and 6% in 940nm image.

show abstract

“…This paper will focus on the effect that using DCT image fusion has on the recognition rate and extends previous work by the authors [7]. Fig.…”

Section: Methodsmentioning

confidence: 67%

Image Fusion based Multi Resolution and Frequency Partition Discrete Cosine Transform for Palm Vein Recognition

Soh

Ibrahim

Abas

et al. 2019

2019 IEEE 6th International Conference on Industrial Engineering and Applications (ICIEA)

Self Cite

View full text Add to dashboard Cite

show abstract

“…To assess the proposed image enhancement method, several palm-vein image recognition systems which use SIFT and RootSIFT features with the CASIA dataset have been implemented. With the SIFT features we applied Euclidian Distance (ED) [6] [8] [9], ED then RANSAC filtering [7].…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…In the 850nm CASIA subset based experiment, 3 images were used as reference images and have used 3 intra class and 6 inter class query images. In [7] the image was first applied with CLAHE and then used histogram stretching on image blocks to enhance the image prior to feature extraction with SIFT. The SIFT features were first matched using the ED then mismatches were estimated with RANSAC algorithm.…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…However, they do not use HE as in [10]. Soh et al [7] proposed an image enhancement process applying a CLAHE operation and then stretching histograms of 8x8 image blocks. A common issue with these techniques is the blocky effect rendered at the edges of the image tiles.…”

Section: Related Workmentioning

confidence: 99%

“…A common issue with these techniques is the blocky effect rendered at the edges of the image tiles. [7] and [11] do not address the blocky effect at all while the 3 pixel overlapping used in [10] is not sufficient to address the blocky effect. Further, using HE can clip out details of the image.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

A Novel Image Enhancement Method for Palm Vein Images

Perera

Khelifi

Belatreche

2022

2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)

View full text Add to dashboard Cite

Palm vein images usually suffer from low contrast due to skin surface scattering the radiance of NIR light and image sensor limitations, hence require employing various techniques to enhance the contrast of the image prior to feature extraction. This paper presents a novel image enhancement method referred to as Multiple Overlapping Tiles (MOT) which adaptively stretches the local contrast of palm vein images using multiple layers of overlapping image tiles. The experiments conducted on the CASIA palm vein image dataset demonstrate that the MOT method retains the finer subspace details of vein images which allows excellent feature detection and matching with SIFT and RootSIFT features. Results on existing palm vein recognition systems demonstrate that the proposed MOT method delivers lower EER values outperforming other existing palm vein image enhancement methods.

show abstract