Detection of heroin covered by skin by using robust principal components analysis

Nowadays, body packing has become an increasingly important approach for trafficking drugs. In this paper, robust principal component analysis (PCA) was demonstrated as a promising suitable method to detect heroin covered by skin based on the profile of energy dispersive X-ray diffraction spectrum. Firstly, some potential impact factors were investigated. Results showed that the signal was optimized when the scatter angle of 8° and the count time of 1 min were chosen. Then, after being truncated, denoised and autoscaled, spectra of heroin, skin and heroin covered by skin were analyzed by robust PCA. Compared with classical counterpart, robust PCA effectively extracted principle components of samples, and successfully identified heroin covered by skin. Finally, based on Robust PCA, a method of identification was set up. This suggests that robust PCA can be employed in detecting drugs packed in body.