太赫兹科学与电子信息学报, 2019, 17 (3): 489, 网络出版: 2019-07-25
高光谱技术在血迹分类识别中的应用
Application of hyperspectral technology in classification and recognition of bloodstains
摘要
利用高光谱技术对血迹种类进行无损识别研究。采用小波变换技术对 400~950 nm之间的原始光谱进行去噪处理,并对处理后的光谱进行特征波段选择,建立全波段和特征波长下的血迹种类识别模型。结果表明,利用特征波长与支持向量机 (SVM)结合建立的血迹种类识别模型的识别准确率及识别时间分别为 98%和 0.2 s,优于全波段建立的模型。研究表明,采用高光谱技术对血迹种类识别是可行的。
Abstract
Nondestructive identification of blood type is studied by hyperspectral technology in this paper. The original spectrum between 400-950 nm is denoised by wavelet transform, and then the characteristic spectrum of the processed spectrum is selected, finally the bloodstain type recognition models in the whole band and characteristic wavelengths are established. The results show that the model established in the characteristic wavelength combined with Support Vector Machine(SVM) can realize recognition accuracy of 98% and recognition time of 0.2 s, which is better than that of the model in full band. Research indicates that the hyperspectral technology is feasible for identifying the bloodstain types.
李成成, 赵明富, 汤斌, 罗彬彬, 邹雪, 王博思. 高光谱技术在血迹分类识别中的应用[J]. 太赫兹科学与电子信息学报, 2019, 17(3): 489. LI Chengcheng, ZHAO Mingfu, TANG Bin, LUO Binbin, ZOU Xue, WANG Bosi. Application of hyperspectral technology in classification and recognition of bloodstains[J]. Journal of terahertz science and electronic information technology, 2019, 17(3): 489.