PENENTUAN KONDlSI PENGUKURAN NIR TERBAIK UNTUK PENENTUAN KANDUNGAN PADATAN TERLARUT BUAH MANGGA SECARA TIDAK MERUSAK MENGGUNAKAN NEAR INFRARED SPECTROSCOPY

ABSTRACTA search for the best NIR experimental condition for nondestructive soluble solids content prediction in mango fruit using near infrared spectroscopy was performed. In this study four treatments that was different in integration time and number of scanning was proposed. A number of 54 mango fruits (Mangifera indica var. Indramayu) were used as samples. The samples were divided into two groups, 27 samples for developing calibration equation and 27 samples for performing validation. Using a portable spectrometer, spectra for each samples was acquired in diffuse reflectance mode at six positions. The SSC of mango was measured using the refraktometer. The Partial Least Squares (PLS) Regression was used to develop a calibration equation. The result showed that the best calibration equetion could be generated from the condition: integration time was 100 ms and number of scanning was 50 scans. Using this condition, the calibration equation resulted in high coefficient of determination (R2 = 0.85), low standard error of calibration, low standard error of prediction and also low bias. Keywords: integration time, number of scanning, calibration equation, soluble solids content; near infrared spectroscopy Diterima: 3 Nopember 2007: Disetujui: 30 Nopember 2007