Self‐geometric relationship filter for efficient SIFT key‐points matching in full and partial palmprint recognition

Abstract

Recently, palmprints have been broadly reported in the literature as an effective biometric modality. Although scaleinvariant feature transform (SIFT)-based features have been proven to be robust against image transformations and deformations, SIFT has not been as successful as other methods in palmprint recognition. In fact, SIFT-based identification has been widely criticised in biometrics due to its high false matching rate. To overcome this weakness, a new filtering method for SIFT-based palmprint matching, called the self-geometric relationship-based filter (SGR-filter) is presented. While existing SIFT matching considers only the relationship between the SIFT points of the query image, on one hand, and their corresponding points in the reference image, on the other hand, SGR-filtering further takes into account the geometric relationship between SIFT points within the query image in comparison with the relationship of the corresponding matched points in the reference image. Assessed with the proposed SGR-filter on various datasets, the SIFT-based palmprint recognition system has been shown to deliver significantly higher performance when compared with the conventional SIFT matching as well as another related key-points filtering technique. Furthermore, experimental results on a number of different full and partial palmprint datasets have shown the superiority of the proposed system over state-of-the-art techniques.

Masked SIFT with align‐based refinement for contactless palmprint recognition

“…For example, on the IITD DB, some methods [30, 32, 35, 36, 38, 47] work on 230 subjects, others [4, 34, 46, 48] work on 235 subjects while [15] works on 215 subjects. Some methods [4, 15, 35, 38, 48] considered the original number of images per subject while others [30, 32, 36, 46, 47] fix this number to five. Some methods [15, 34, 35, 38, 47, 48] considered a training set with a different number of images while others [4, 30, 32, 36, 46] do not.…”

“…These matches are further refined by comparing the local palmprint descriptors at their locations and exclude the mismatched ones. The authors in [38] apply a self-geometric relationship-based filter (SGR-filter) to exclude the false matches. It takes into account the geometric relationship between SIFT points within the query image in comparison with the corresponding ones in the target image.…”

“…Among the descriptor‐based methods, the SIFT‐based ones prove robustness against geometric transformations and give better results over different contactless databases [4, 34, 38]. The SIFT keypoints are extracted and described from query and target palms, then matched using the original Lowe's method [39].…”

Rotation and Translation Invariant Palmprint Recognition With Biologically Inspired Transform

“…This method can reduce the sensitivity due to the limited size of the training data. Almaghtuf and Khelifi [46] presented a SIFT-based palmprint matching method, which can take into account the geometric relationship between SIFT points within the query image in comparison with the relationship of the corresponding matched points in the reference image.…”

Combining multi‐wavelet and CNN for palmprint recognition against noise and misalignment

“…Personal identity is recognised by extracting the palmprint texture feature between the end of the finger and the wrist, which is unique, stable, and reliable [1,2]. Traditional palmprint feature extraction methods are mainly divided into four types based on structural feature [3,4], statistical feature [5][6][7][8][9], subspace feature [10][11][12][13][14][15][16], and coded feature [8,[17][18][19][20][21]. The classification methods basically used the neural network and SVM classifiers.…”