Research On Blood Stain Detection Based On Hyper-spectral Imaging Technology

The most common trace of evidence at a crime scene is blood.Bloodstain,as one of most important material evidences in criminal proceedings,has a high probability of occurrence in criminal events,and has a strong objective stability of information expression.At the same time,it is also vulnerable to the influence of various external environmental factors,and is extremely complex.In judicial scientific proof,it is regarded as one of the most important material evidence.Through the available feasible technical means to deal with the blood at the scene of the crime,a large number of potential material evidence information related to the case can be obtained.The identification and detection of blood and prediction of the time of blood formation have always been very important steps in the detection of criminal cases.At present,the traditional blood detection in forensic work is usually direct observation method,reagent method,etc.The photo-electricity technology has also been used in blood analysis,but its application limitation is large,and the application universality is not high.These methods hinder the application of these methods in practical cases,and many of these methods will destroy the components of blood,which is not conducive to the acquisition of material evidence and the further investigation of the case.In recent years,hyper-spectral imaging technology has gradually developed and matured.Based on its unique three-dimensional data structure and the characteristics of "atlas unity",the technology has been applied in more and more scientific fields,from aerospace,ocean detection to medical imaging.Food safety monitoring and so on,in these extensive applications,the strong adaptability of its function has gradually emerged,so it has been paid close attention to more and more countries.Aiming at the forensic blood detection in criminal cases,especially the detection of potential blood at the scene of the crime and the prediction and inference of the formation time of blood,a method based on hyper-spectral imaging is proposed to detect the spectral characteristics of blood.The proposed method can effectively identify and detect blood samples and accurately predict the time.The hyper-spectral imaging technology is successfully combined with forensic criminal detection,and the new application of hyper-spectral imaging technology is realized.Therefore,the research of blood spectral feature detection method based on hyper-spectral imaging technology has great research space and application prospect.The main work of this study is as follows:(1)A new method of blood trace recognition and detection is proposed,which is based on hyper-spectral imaging technology to detect spectral inequalities of blood trace.The hyper-spectral imaging system platform is adjusted to acquire the hyper-spectral three-dimensional data images of the prepared blood stains and suspected substances through hyper-spectral imaging technology,and the band inequality model of blood stains based on spectral feature transformation is proposed through processing and analysis.The proposed model is used to realize the correct recognition and extraction of blood stains in various stains on textile bases.(2)The data acquired by hyper-spectral imaging technology can simultaneously possess the color change information of blood-stain spatial image and the spectral reflectance intensity distribution information of corresponding pixels of blood-stain.The spectral reflectance intensity distribution of blood-stain at different time on white writing paper is analyzed,and a new idea of predicting and inferring the formation time of blood-stain is put forward,which is used to predict the time of blood-stain.The model is based on the oxidative denaturation of hemoglobin in blood in vitro.The three-dimensional blood trace data cube obtained by hyper-spectral imaging spectrometer is used as the technical support.The blood trace spectrum and the corresponding time data are extracted for statistical analysis.The model is assisted by the specific calculation process such as the non-linear iteration algorithm,the non-linear regression algorithm and the canonical correlation analysis algorithm.A model formula which can fully express the correlation between blood trace spectrum and its time is obtained.The main function of this model is to effectively predict and infer the formation time of blood stains.The experimental data show that the model can accurately predict the formation time of blood stains,and has great practical application prospects.