Response Characters And Detection Of Near-infrared Spectroscopy On β-glucan In Oat Grains

In recent years,people have paid more and more attention to their health,and their eating habits have also changed.Small crop has become an indispensable food in people’s lives.Oat is a small crop with rich and multiple functional ingredients.Oat yield ranks sixth among all grain crops.β-glucan is considered to be one of the most potential functional components in oats.Therefore,it is great significance to detect quickly and accurately β-glucan content for evaluating quality and function of oats.At present,the determination method of oat β-glucan content has the disadvantages of low efficiency,cumbersome operation and expensive cost,which is difficult to meet the needs of future large-scale oat production and industrialization development.Near-infrared spectroscopy technology is a method with the character of fast,non-destructive,accurate and efficient,and has been widely used in agriculture and other fields.Therefore,detecting the β-glucan content in oat grains based on the near-infrared spectroscopy has important practical needs and application value.In this study,63 grains of oat germplasm were used as the experimental material to measure theβ-glucan content and their near-infrared spectrum.After pretreatmenting the original spectral data（16mathematical transformation methods and 4 correction methods）,the response characteristics of near-infrared spectroscopy to oat grain β-glucan were analyzed combined with three types of spectral models.Then,the characteristic of near-infrared spectral of β-glucan in oat grain was explored and determined.The detection model of near infrared spectroscopy was tried to construct.The aim was to clarify the quantitative relationship betweenβ-glucan in oat grain and near infrared spectroscopy,and to study the effect of pretreatment methods on detection of near-infrared spectroscopy on β-glucan in oat grains.The main conclusions were as follows:（1）The near-infrared absorption decreased with the increase of β-glucan content in oat grain.There was a negative correlation between β-glucan content and near-infrared absorption.In addition,the characteristic area of the near-infrared spectra expressing β-glucan content were 9989 cm^-19777cm^-1,9300cm^-18800 cm^-1,7594 cm^-16992 cm^-1 and 4000 cm^-1.The 8840.091 cm^-1,7324.316 cm^-1,6992.62 cm^-1,8816.949 cm^-1,8469.825 cm^-1,8890.231 cm^-1,9075.364 cm^-1,9391.633 cm^-1,8867.09 cm^-1,9507.341 cm^-1and 4512.612 cm^-1 were proved to related closely to β-glucan content,which showed it was feasible to detect rapidly and accurately β-glucan content in oat grain using near infrared spectroscopy.（2）The established MLR prediction model performed best（correction model R²=0.788,RMSE=0.556,RPD=2.196,verification model R²=0.842,RMSE=0.545,RPD=2.453）based on the mathematical transformation method R²,which could achieve the accurate detection on β-glucan content in oat grain using near-infrared spectroscopy technology.In addition,the quantitative relationship betweenβ-glucan content in oat grain and near-infrared spectrum could be described using the MLR spectral characteristics based on R² pretreatment conditions.（3）The pretreatment method has an effect on the detection of β-glucan content in oat grain using near-infrared spectroscopy.The square pretreatment of the original spectrum（mathematical transformation method 15）could significantly increase the detection accuracy of near-infrared spectrum of on β-glucan content.In addition,MSC correction processing generally performed better in various models,followed by differential processing.Comparing the mathematical transformation processing and correction processing,it was showed that the correction processing had a better effect in improving the performance of the model.