基于分形维数差模法的荧光光谱识别  

Study of Fluorescence Spectrum Identification Based on the Difference Module of the Fractal Dimensions

作　　者：樊迪[1] 王乐新[2] 杨蔚然[1] 谭可丽 邱润泽 

机构地区：[1]黑龙江八一农垦大学信息技术学院,大庆163319 [2]黑龙江八一农垦大学理学院

出　　处：《黑龙江八一农垦大学学报》2016年第4期130-134,共5页journal of heilongjiang bayi agricultural university

基　　金：黑龙江省自然基金资助项目(F201427);黑龙江省教育厅资助项目(10541155;12521376);黑龙江省农垦总局科技项目(HNK11A-06-09);大庆市科技局资助项目(SGG2008-041);大学生省级创新创业训练项目(201410223009)

摘　　要：提出一种基于分形理论的光谱分形特征识别光谱曲线的分析方法,选取50个待测血清样品,分别测量血清样品在波长为260、290、350和580 nm激发光下产生的荧光光谱。应用分形理论计算光谱曲线的分形维数,利用分形维数的差模识别不同血清(正常、高甘油三脂、高胆固醇、高血糖)的荧光光谱。结果表明,高甘油三脂血清光谱识别率88%,高胆固醇血清识别率81%,正常血清识别率为75%,高血糖血清识别率为60%,为分形理论在光谱识别上应用作了初步探索。The fractal feature of the spectrum based on the fractal theory was proposed to identify the spectral curves. The fluorescence spectra of 50 serum samples were measured by using the exciting light with the wavelength of 260,290,350, and 580 nm, respectively. The fractal dimensions of the spectral curves were calculated by the fractal theory. The difference module of the fractal dimensions was used to identify the fluorescence spectra of normal, high blood lipid, high cholesterol, and high blood glucose serum. The results indicated that the recognition rate was 75%,88%, 81% and 60%, respectively. The research showed a preliminary study for the fractal theory in spectral identification.

关 键 词：分形理论 光谱识别 差模 血清 

分 类 号：O433.4[机械工程—光学工程]