HYPERSPECTRAL IMAGING SYSTEM IN FORENSIC SCIENCE: DEFINITION OF THE TEST BENCH AND AGE ESTIMATION OF BLOODSTAINS ON DIFFERENT SUBSTRATES

In the last years hyperspectral imaging system has been used in different fields, and recently it has been taken in consideration as a potential tool for forensic science. Hyperspectral imaging system is both an imaging and spectroscopy technique and is considered to be useful in forensic science as it is a fast, portable, and non-destructive technique, able to maintain the integrity of the biological traces found at the crime scene. In general, several traces can be found at the crime scene; the most important is blood since several information can be obtained from it. In fact, visual inspection of the bloodstains gives information about the motion of the victims and the suspects, the DNA analysis performed on the stain is useful for the identification of the suspects, and the determination of the age of the bloodstains gives insight about the temporal aspect of the crime. Haemoglobin spectrum dominates the visible spectrum of a bloodstain, and it is the main blood chromophore. When haemoglobin is found outside the human body, i.e., in a bloodstain, it undergoes a series of degradation processes which lead to the formation of the haemoglobin derivatives. The changes in the haemoglobin and haemoglobin derivatives concentration lead to the colour change of the bloodstains, from red to dark brown, according to the fact that the derivatives are characterised by different absorption spectra. Consequently, the bloodstain spectrum changes over time, hence the temporal analysis of the bloodstain spectrum can be used for age estimation purposes since the temporal behaviour of the spectrum relates to the chemical changes occurring in the bloodstain. In this study, hyperspectral imaging system has been used to perform the temporal analysis of bloodstains deposited on different substrates to determine a set of temporal parameters related to the age of the stain. Moreover, since the characteristics of the illumination system affect the performance and reliability of the hyperspectral imaging system, the test bench has been defined to obtain the spectral data for the temporal analysis. Subsequently, the curves, expressing the relation between the age of the bloodstain on different substrates and the temporal parameters defined, have been determined. After the correction of the bloodstain spectra, which are influenced by the substrate colour and texture, via a Neural Network model, it has been possible to obtain a general curve for each defined temporal parameters with optimal RMSE and R2 values which could be used to perform the age estimation of the bloodstain when the substrate is not known.