Quantitative Evaluation of Fingerprint Evidence Value Based on Python
10.19756/j.issn.0253-3820.241396
- VernacularTitle:基于Python编程语言的手印证据价值的定量评估
- Author:
Zhi-Ze XU
1
;
Meng WANG
;
Rong-Wei MA
;
Jie LI
;
Ming LI
;
Chuan-Jun YUAN
Author Information
1. 中国刑事警察学院刑事科学技术学院,沈阳 110035
- Keywords:
Fingerprint development;
Fingerprint identification;
Nanomaterials;
Deep learning;
Evidence value
- From:
Chinese Journal of Analytical Chemistry
2025;53(4):590-601,中插12-中插22
- CountryChina
- Language:Chinese
-
Abstract:
A deep learning-based method for recognizing the minutiae in fingerprint,as well as a Python programming-based evaluation system for quantifying the evidence value of fingerprint was proposed.Firstly,latent fingerprints,which were developed using a series of fluorescent nanomaterials synthesized by chemical methods,were used as unknown fingerprint(UKFP),while ink impressed fingerprints were used as known fingerprint(KFP).Then,the bifurcations and terminations in minutiae were recognized using the improved YOLOv8 deep learning model.After that,the similarity index(Sim.)of UKFP vs KFP were calculated by analyzing the angle similarity factor(α)and the curve similarity factor(β)between UKFP and KFP,meanwhile,the sensitivity index(Sen.)were calculated by analyzing the fineness factor(γ)between UKFP and KFP.The evidence value(EV)of fingerprint was thus obtained by the combination of Sim.and Sen..The calculation formulas for above evaluation factors(i.e.α,β and γ),evaluation indexes(i.e.Sim.and Sen.),and EV were also put forward.Finally,the evaluation system for quantifying the evidence value of fingerprint was established,the feasibility and reliability of this system were verified,and the external factors that impacted on Sim.,Sen.,and EV were investigated in detail.The Python-based evaluation system for quantifying the evidence value of fingerprint could achieve the goals objectively,comprehensively,accurately and efficiently,exhibiting easy operability,high efficiency,responsiveness and reliability.This research was expected to provide beneficial references for quantitatively evaluating and thoroughly developing the evidence value.
