利用可見光／近紅外線光譜判定絞碎豬小里脊肉之新鮮度   全文下載

Evaluation of Freshness of Ground Pork Tenderloin by Visible/Near Infrared Spectroscopy

本研究將 120 件絞碎台灣黑毛豬小里脊肉樣本儲存在7°C共5天，每天以蒸餾水萃取，量測其萃取液的可見光／近紅外線光譜（600-1368 nm），並配合 pH 值、揮發性鹽基態氮（VBN）和生菌數（APC）三個新鮮度指標，建立絞碎小里脊肉樣本之新鮮度預測模式。結果顯示在 600-840 nm 範圍之光譜，最有利於新鮮度預測模式之建立，再以此範圍內四個特定波長（608, 624, 656 和 732 nm）之吸收值、搭配鮮度指標作判定。判定則採用部分最小平方迴歸（PLSR）法，以虛擬迴歸技術去除第三天數據後，對 200 組樣本之判定準確率可達90%。"

Visible/near-infrared transmittance spectra (600-1368 nm) of the aqueous extracts of 120 ground pork tenderloin samples from Taiwan black-hair hogs were used to correlate with the freshness-relating data including pH, volatile basic nitrogen and aerobic plate count during 7°C storage for 5 days. The spectra in 600-840 nm regions varied during storage with the data highly correlated with the freshness parameters. Four characteristic wavelengths (608, 624, 656 and 732 nm) were selected for freshness evaluation based on partial least square regression (PLSR) model using different freshness criteria for dummy regression. The accuracies of freshness evaluation can reach 90% with the PLSR models excluding the data in Day 3.