基于分子光谱分析的人指甲无损鉴别及性别刻画

Nondestructive Identification and Gender Characterization of Human Nails Based on Molecular Spectroscopy Analysis

指甲等人体生物组织的鉴定在刑事案件侦查中发挥着重要作用。为了对犯罪现场提取的指甲组织进行快速无损鉴别,提出了一种基于分子光谱分析和机器学习的人指甲无损鉴别和性别刻画方法。通过采集120个同年龄段不同性别人指甲样本的红外光谱数据,建立了多种分类预测模型。借助主成分分析技术降维提取3个主成分,对样本进行交互验证,并对比了Fisher判别函数、多层感知器及反向传播(BP)神经网络模型的识别效果。实验结果表明:多层感知器模型的分类识别率可达到91.4%,优于Fisher判别分析模型;基于粒子群优化算法的BP神经网络模型分类效果最佳,识别率达到97.7%。

The inspection and identification of human biological tissues,such as nails,play an essential role in investigating several criminal cases.To quickly and nondestructively identify nail tissues extracted from crime scenes,this paper proposes a nondestructive identification and gender characterization method of human nails based on molecular spectroscopic analysis and machine learning.We establish various classification prediction models by collecting 120 infrared spectroscopy data of different gender nail samples of the same age group.Using principal component analysis technology,dimensionality reduction is used to extract 3 principal components,and the samples are interactively verified.The recognition effects of Fisher discriminant function,multilayer perceptron,and back propagation(BP)neural network model are also compared.The experimental results show that the classification and recognition rate of the multilayer perceptron model can reach 91.4%,which is better than the Fisher discriminant analysis model.The BP neural network model based on the particle swarm optimization algorithm has the best classification effect,with a recognition rate of 97.7%.