Spectrometry Evaluation of B10 Averrhoa carambola L. and Sala Mango Physical Attributes

摘要

The physical quality attributes of B10 carambola and Sala mango are normally graded based on its colour index and size (weight). For colour classification, the indexing process is conducted by trained staffs from Federal Agricultural Marketing Authority (FAMA). However, this method is highly subject to error since the evaluation is made through human visual perception. This paper presents an innovative application of visible and near infrared (NIR) spectroscopy in quantitative measurement of carambola and mango physical quality, two of the Malaysian most popular tropical fruits. For comparative analysis, the spectroscopy measurement was conducted using reflectance and interactance techniques. The best algorithm for predicting carambola and mango index has been generated by MLR + SG (reflectance: R~2 = 0.934; RMSEP = 0.489) and MLR + MSC (reflectance: R~2 = 0.931; RMSEP = 0.504) respectively. On the other hand, the best algorithm for predicting carambola and mango weight has been generated by MLR + SG (interactance: R~2 = 0.754; RMSEP = 17.454 g) and MLR + SG (reflectance: R~2 = 0.769; RMSEP = 89.122 g) respectively.