Near infrared spectrum-based fiber component identification method and module

Abstract

The invention discloses a near infrared spectrum-based fiber component identification method and module. The near infrared spectrum-based fiber component identification method comprises the steps of a) establishing an identification model, b) pre-processing before classification and c) performing a classification and identification algorithm. The near infrared spectrum-based fiber component identification module comprises a fiber component identification system software, the fiber component identification system software adopts the near infrared spectrum-based fiber component identification method and is compiled in a C# language through Visual Studio 2017. According to the present invention, the spectral data, collected by a portable near-infrared spectrometer, of a sample, is preprocessed through a first-order derivative, classification and discrimination are conducted through the established PLS-DA model, and the identification of fibers, such as cotton, wool, silk, aramid fibers etc., can be achieved; the C# language is adopted to realize the development of a fiber component identification system on a program, and through the combination with the portable near-infrared spectrometer, the rapid, online and nondestructive identification of a large class of fiber components is realized.

Description

【Technical field】 [0001] The invention relates to the technical field of fiber identification, in particular to the technical field of a fiber component identification method and module based on near-infrared spectroscopy. 【Background technique】 [0002] When using the spectrum to identify the fiber composition, the spectral band range usually used to identify the composition of the textile material using the spectral information of the textile material is the mid-infrared band (2500nm-25000nm) in the infrared band. Using the mid-infrared band to identify and analyze the composition of textile materials has a high identification accuracy, but due to the test principle, the structure of the test instrument is relatively complex, there are many components, the size and weight of the instrument are large, it is difficult to achieve portability, and it is difficult to realize the test area. of imaging. [0003] The use of near-infrared spectroscopy to identify the composition o...

AI Technical Summary

Problems solved by technology

When using spectra to identify textile materials, a relatively simple method is to compare the original spectral curves with the naked eye, analyze the differences in the spectral curves of different samples, and judge the type of sample to be tested accordingly, although this method can achieve textile material fiber Identification of components, but there are problems such as being greatly affected by personal subjective factors and low identification efficiency

In addition, the near-infrared band spectra of textile materials with different components have the problems of weak absorption peak intensity, spectral bandwidth and characteristic peak overlap.

Images