Establishment Of Near Infrared Spectral Model Of Sugar Content In Sweet Corn

Sweetcorn is one of the most important dual-purpose fruit and vegetable crop in the world.With the further improvement of people's living standard,higher quality of sweetcorn is required.More and more attention are paid on the genetic improvement of sweetcorn quality.At the present,the bottleneck of sweetcorn breeding in south China is the lack of high quality varieties,such as fruit varieties.The key to solve this problem is to quantify the quality indices.Therefore we need to establish a quick and easy identification evaluation system and select high-quality resources for genetic improvement.One of the most important aspects of the genetic improvement of sweetcorn is to breed some endosperm mutant strains which can produce high level sugar content.First this study compares the applicability of 3,5-dinitrosalicylic acid method against enzymatic method in determination for sweet corn sugar content.Then we chose the better one to determine one hundred and four sweetcorn sugar content,using near-infrared spectrum analysis to collect spectral information of these samples at the same time.The modified Partial Least Squares regression,spectral preprocessing and mathematical treatment were employed to analyze the full spectrum segment.Models were built on total polysaccharides,starch,glucose,fructose and sucrose content on this research.The results showed as fallows:1.From the methodological comparison of enzymatic method and 3,5-dinitrosalicylic acid method for the determination of sugar content of sweet corn,the precision and stability of the 3,5-dinitrosalicylic acid method is 0.994%,0.702%,which smaller than enzymatic method(3.924%,2.298%),and the average recovery rate relative to standard deviation is 1.031% higher than enzymatic method(0.777%).From the results analysis can be seen,both methods can be used to determine the content of sweet corn sugar,but result of 3,5-dinitrosalicylic acid method is higher.In order to obtain the accurate chemical value of sweet corn and a prediction model of sugar index,the enzymatic method was used to detect the chemical value of sweetcorn for near-infrared spectroscopy.2.It was six of the best combination of Derivative Treatment Gap and Smoothing in the the content of total polysaccharide,glucose,fructose and sucrose of sweet corn.But the combination of Derivative Treatment Gap and Smoothing at two were better for the starch content.The best spectral treatment of total polysaccharides,fructose and sucrose content was Standard Normal Variant plus Detrend,and the optimal spectral treatment ofstarch and glucose was Standard Multivariate Scatter Correction and Weighted Multivariate Scatter Correction.The optimal derivative condition of total polysaccharides,starch,glucose,fructose and sucrose content was 3,2,2,1,3,respectively.External validation of the calibration models of the five sweetcorn qualitiy contents was used,the correlation coefficients were 0.834,0.783,0.593,0.780,0.891,respectively.In addition to the low correlation coefficient of glucose prediction,the other four indicators have reached more than 0.780,which can get better prediction results.It means that the model can be used to determine the sugar content in the complete dry grain of sweet corn.