Dimensionality reduction and visualisation of hyperspectral ink data using t-SNE

•The t-SNE algorithm is introduced into forensic ink data analysis.•Created hyperspectra database of inks from 60 pens, from different manufactures, type and colour.•Compared the clustering quality of t-SNE against PCA on hyperspectral ink data.•Clustering quality compared using four different clust...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Forensic science international Ročník 311; s. 110194
Hlavní autoři: Melit Devassy, Binu, George, Sony
Médium: Journal Article
Jazyk:angličtina
Vydáno: Ireland Elsevier B.V 01.06.2020
Elsevier Limited
Témata:
Algorithms
Cameras
Clustering
Cultural heritage
Data processing
Datasets
Dimensionality reduction
Discriminant analysis
Embedding
Feature extraction
Forensic science
Forensic sciences
hyperspectral imagery
Hyperspectral imaging
information processing
Ink analysis
Narrowband
Principal components analysis
Probability
quantitative analysis
Reduction
Software
Spectrum analysis
t-SNE
Visualisation
Visualization
Dimensionality reduction
Visualisation
t-SNE
Ink analysis
Hyperspectral imaging
ISSN:0379-0738, 1872-6283, 1872-6283
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:•The t-SNE algorithm is introduced into forensic ink data analysis.•Created hyperspectra database of inks from 60 pens, from different manufactures, type and colour.•Compared the clustering quality of t-SNE against PCA on hyperspectral ink data.•Clustering quality compared using four different clustering quality indexes.•The t-SNE provided better visualization and clustering score. Ink analysis is an important tool in forensic science and document analysis. Hyperspectral imaging (HSI) captures large number of narrowband images across the electromagnetic spectrum. HSI is one of the non-invasive tools used in forensic document analysis, especially for ink analysis. The substantial information from multiple bands in HSI images empowers us to make non-destructive diagnosis and identification of forensic evidence in questioned documents. The presence of numerous band information in HSI data makes processing and storing becomes a computationally challenging task. Therefore, dimensionality reduction and visualization play a vital role in HSI data processing to achieve efficient processing and effortless understanding of the data. In this paper, an advanced approach known as t-Distributed Stochastic Neighbor embedding (t-SNE) algorithm is introduced into the ink analysis problem. t-SNE extracts the non-linear similarity features between spectra to scale them into a lower dimension. This capability of the t-SNE algorithm for ink spectral data is verified visually and quantitatively, the two-dimensional data generated by the t-SNE showed a better visualization and a greater improvement in clustering quality in comparison with Principal Component Analysis (PCA).
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:0379-0738
1872-6283
1872-6283
DOI:10.1016/j.forsciint.2020.110194