Application of NIR spectroscopy for the prediction of soluble solids in guava pulp

Acta Agronómica

Ibarra Espinosa, Dora Mónica¹  Muñoz Flórez, Jaime Eduardo¹  Gutiérrez Devia, ányelo Andrey¹  Belalcázar Martínez, John Eiver²  Caetano, Creuci María¹  Caicedo Arana, álvaro³ 

[1] Universidad Nacional de Colombia, Palmira, Colombia;CIAT, Palmira;Corporación Colombiana de Investigación Agropecuaria - Corpoica, Palmira, Colombia

关键词: Near infrared spectroscopy;    °brix;    calibration;    guavaIntroducción El guayabo (Psidium guajava L;   

DOI  :  10.15446/acag.v64n2.40107

学科分类：农业科学（综合）

来源: Universidad Nacional de Colombia * Facultad de Ciencias Agropecuarias Palmira

【 摘 要 】

Prediction of soluble solids in pulp of 92 samples of guava (Psidium guajava L.) and (Psidium friedrichsthalianum) in two maturity stages was made by using NIRS technique (near infrared reflectance spectroscopy). Each fruit was processed to obtain the edible fraction in which it was determined the soluble solids content (%). Two subsamples were taken and scanned in NIR spectrophotometer at the range of 400 to 2500 nm. Calibration regression models were generated by using MPLS (modified partial least squares) at the range of 1108 - 2498.2 nm. For the choice of the final model of the NIR calibration the criteria considered were the coefficient of determination (R2) and the residual predictive deviation (RPD). Analysis of variance and mean test employing the statistical program SAS v 9.0 were realized in order to establish differences in the results obtained by the primary or reference method and NIR. The results showed a high predictive power of the model (R2 = 0.990) when compared the primary and NIR measurements. The RPD value was 6.20, indicating excellent prediction accuracy. Primary estimates and NIR not showed significant differences. Ho wever there were significant differences at level of samples.

【 授权许可】

Unknown