Research And Construction Near Infrared Spectrum Analysis Software For Ocean Optics Spectrometer

With the continuous development of various detection techniques,the nondestructive and rapid detection method has become an inevitable trend in the future.NIR technology has been developing rapidly.It has been widely used in agriculture,pharmaceutical industry,fruit quality inspection and other related fields because of its fast detection,nondestructive testing and easy operation.This technology includes three steps:the acquisition of the spectrum,the management of the spectral data,and the analysis of the spectral data.The software package of Ocean Optics near-infrared spectrometer only has the function of spectral acquisition,while managing and analyzing data needs the help of Matlab and other tools,which is complex and inefficient.In order to solve this problem,this paper uses object-oriented programming idea to develop a spectral management and analysis software for this spectrometer.Using the software can easily manage the data collected by the spectrometer,at the same time,the software integrates the commonly used algorithms in the field of near-infrared spectroscopy analysis,including filtering,smoothing and other preprocessing methods,as well as PCA,PLS,SVM and other model building algorithms.Users can make use of these algorithms to identify and predict the unknown samples by modeling the known samples.Using the spectral data of three varieties of corn seeds at the present stage to test the effectiveness of the system modeling,the established model can identify the variety information well,and the model effect is ideal,which lays the foundation for the further extension of the system to the identification of more types of materials.