Skull similarity comparison based on SPCA

Abstract

As an important biological feature of the human, the skull plays an active role in assisting criminal investigation, victim identification, etc. This paper proposes a method based on Sparse Principal Component Analysis (SPCA) for comparison of skull similarity. Compared with Principal Component Analysis (PCA), SPCA can not only effectively reduce the data dimension,but also produce sparse principal components which are easy to explain. Each principal component of PCA is a linear combination of all original variables. It’s difficulty in explaining the corresponding relationship between principal components and features. SPCA makes the loadings sparse, and thus highlights the main part of the principal component, which can solve the problem of PCA that has difficulty in explaining the result. The experimental results show that the dimensionality reduction data by SPCA is superior to PCA in the aspects of complexity, discrimination, stability, interpretability, and similarity evaluation. These indicate that the comparison of skull similarity based on SPCA is accurate and stable, which can provide an effective direction for improving the accuracy of craniofacial reconstruction and obtaining accurate reconstruction results.