Related Experiment Video
Updated: Nov 24, 2025

Measuring the Complete-arch Distortion of an Optical Dental Impression
Published on: May 30, 2019
Fingerprint restoration using cubic Bezier curve
Yanglin Tu1, Zengwei Yao2, Jiao Xu1
1Zhuhai People's Hospital (Zhuhai Hospital Affiliated with Jinan University), No. 79, Kangning Road,Xiangzhou District, Zhuhai , 519000, Guangdong, China.
This study introduces a new algorithm using Bezier curves to restore missing fingerprint ridges, significantly improving matching accuracy for incomplete biometric data. The method enhances authentication reliability by effectively reconstructing damaged fingerprint images.
Area of Science:
- Biometrics
- Computer Vision
- Pattern Recognition
Background:
- Fingerprint biometrics are crucial for authentication.
- Matching incomplete fingerprints with missing minutiae or ridges presents a significant challenge.
- Many fingerprint matching attempts fail due to image incompleteness.
Purpose of the Study:
- To develop a novel algorithm for detecting and restoring fragmented ridges in incomplete fingerprints.
- To represent fingerprint data using Bezier curves for efficient data handling.
- To improve the accuracy of fingerprint matching with damaged or incomplete images.
Main Methods:
- Fingerprints are modeled using Bezier curves, with control points representing fingerprint fragments.
- A novel algorithm is proposed to detect and restore fragmented ridges.
- The proposed method's performance is evaluated on synthetic (SFinGe) and real (FVC2004 DB1) fingerprint datasets.
- The algorithm's effectiveness is compared against convolutional neural network models (FDP-M-net, U-finger).
Main Results:
- Bezier curve representation reduces data size by 89% with lossless restoration.
- Fingerprint matching scores increased by 39.54% on synthetic data and 13.22% on real data.
- Equal Error Rate (ERR) and False Match Rate (FMR1000) were significantly reduced on both datasets.
- The proposed algorithm outperformed FDP-M-net and U-finger in improving matching scores.
Conclusions:
- The algorithm successfully repairs and reconstructs ridges in damaged fingerprint images.
- The proposed method enhances the accuracy of fingerprint matching.
- This contributes to more reliable authentication systems using incomplete biometric data.
More Related Videos
Related Concept Videos
Reconstruction of Signal using Interpolation
Elevation of Intermediate Points on Vertical Curves
IR Frequency Region: Fingerprint Region
Deformations in a Symmetric Member in Bending
When the member is segmented into tiny cubic elements, it is observed that the primary stress...
Tangent to a Curve
Vertical Curve: Problem Solving

