A Filtering Method for SIFT based Palm Vein Recognition

2022-11-30

International Conference on Digital Image Computing: Techniques and Applications

AI Summary

In palm vein recognition, minor hand pose variations induce skin deformation, causing substantial intra-class variability in vein patterns and severely degrading the robustness of SIFT-based feature matching. To address this, we propose Median Distance (MMD) filtering—a novel post-processing technique for SIFT matches. MMD is the first method to incorporate median statistics into SIFT matching refinement: it jointly analyzes deviations of keypoint coordinates in both x- and y-dimensions from their respective means and medians, and employs a dual-threshold criterion coupled with a rule-driven mechanism to precisely identify and remove outliers that conventional methods fail to detect. Importantly, MMD requires no model retraining and seamlessly integrates into existing SIFT pipelines. Evaluated on the CASIA 850 nm dataset, our approach achieves a significantly lower Equal Error Rate (EER) than state-of-the-art SIFT-based methods, demonstrating superior robustness against deformation-induced distortions and practical applicability.

Technology Category

• Computer Vision: Motion & Tracking
• Machine Learning: Dimensionality Reduction/Feature Selection
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

A key issue with palm vein images is that slight movements of fingers and the thumb or changes in the hand pose can stretch the skin in different areas and alter the vein patterns. This can produce palm vein images with an infinite number of variations for a given subject. This paper presents a novel filtering method for SIFT based feature matching referred to as the Median Distance (MMD) Filter, which checks the difference of keypoint coordinates and calculates the mean and the median in each direction, and uses a set of rules to determine the correct matches. Our experiments conducted with the 850nm subset of the CASIA dataset show that the MMD filter can detect and filter false positives that were not detected by other filtering methods. Comparison against existing SIFT based palm vein recognition systems demonstrates that the proposed MMD filter produces excellent performance with lower EER values.

Problem

• Addresses palm vein pattern variations due to hand movements.
• Introduces MMD filter to improve SIFT-based feature matching accuracy.
• Reduces false positives and achieves lower Equal Error Rate (EER).

Innovation

• MMD Filter improves SIFT-based feature matching
• MMD Filter reduces false positives in vein recognition
• Achieves lower EER in palm vein recognition