摘要
近年来法医个体识别的实践工作中,主流方法仍然是指纹、DNA、齿科对比等,但是在一些极端情况,例如焚烧、腐败、白骨化等情况时,常规方法却难以奏效。基于此,提出基于蝶窦的三维点云深度学习身份认定方法,进行训练和测试,效果证明该思路的可行性。
In the practical work of forensic individual identification in recent years,mainstream methods are still fingerprints,DNA,dental comparison,etc.However,in some extreme situations,such as incineration,rottenness,and ossification,conventional methods do not work.Based on this situation,a 3 D point cloud deep learning identification method based on sphenoid sinus was proposed,and training and testing were carried out.The results proved the feasibility of this idea.
作者
文瀚杰
WEN Hanjie(Nation Key Laboratory of Fundamental Science on Synthetic Vision,Sichuan University,Chengdu 610065)
出处
《现代计算机》
2021年第13期86-89,共4页
Modern Computer
基金
成都市科技局科研项目(No.2018YF0500069SN)
四川省科技厅科研项目(No.2019YFH0193)。
关键词
法医
点云
蝶窦
深度学习
身份认定
Forensic
Point Cloud
Sphenoid Sinus
Deep Learning
Identification
