Please use this identifier to cite or link to this item:
標題: 以近紅外線光譜檢測稻米之食味
Rice Taste Evaluation Using Near-Infrared Spectrum
作者: 張明暉
Chang, Ming-Hui
關鍵字: Near infrared
出版社: 生物產業機電工程學系
摘要: Abstract Rice moisture and protein contents affect its sensory taste quality. This study constructs calibration curves of sensory taste of boiled rice measured using panel tests and AN800 machine to correlate with moisture and protein contents of the rice measured using near infrared (NIR) reflexed spectra analysis. The developed curves can be used for non-destructive and rapid inspection of rice taste quality. The rice samples for moisture content calibration were Taikeng-9 rice direct purchased in market and then adjusted to different moisture contents from 10-16% (w.b.). The rice samples for protein measurement were Taikeng-9 and Tainung-71 which were grown in paddy field and applied fertilizer with different nitrogenous contents of 0, 40, 80, 120, 160, 200 kg/ha, consequently to have different protein content. The accuracy of rice (Taikeng-9) moisture content measurement using the AN800 machine was compared with oven method, which relationships between the two methods for brown rice was mAB= 0.878mOB + 2.243 and for milled rice was mAM=1.109mOM -0.117, with correlation coefficient R2 =0.987 and 0.979 respectively. The relationships of rice protein content measurement using AN800 machine and Kjetec 2400 instrument for brown rice (Tainung-71) was pAB=0.775pKB+1.516 and milled rice (Taikeng-9) was pAM =1.027pKM- 0.289, with correlation coefficient R2 = 0.916 and 0.979 respectively. The Taikeng-9 rice (brown and milled) calibration curves using NIR reflexed spectra at wavelengths 700-2500 nm and MPLS regression correspondent with AN800 measurement have R2 for moisture content of (0.997 and 0.999), for protein content of (0.955 and 0.967) and for sensory taste of (0.990 and 0.969), respectively, with SEC (0.090 and 0.056), (0.094 and 0.109) and (0.655 and 1.148), respectively, and SEP (0.111 and 0.088), (0.148 and 0.162) and (1.554 and 1.270), respectively. The panel test results indicated that the moisture content of Taikeng-9 milled rice has positive correlation with appearance, cohesiveness and overall flavor scores of boiled rice, however has negative correlation with hardness of the boiled rice; protein content of milled rice has negative correlation with taste, cohesiveness and overall flavor, however, has positive correlation with hardness. The relationships of the overall flavor scores from the panel tests and taste scores from the AN800 machine of milled rice corrected to moisture content (10-15%, w.b.) was TMT = 1.292TMA3+1.615TMA2 +1.341TMA -0.068 and corrected to protein content (4.9-6.6%, d.b.) was TPT =0.330TPA3 -0.027TPA2+0.206TPA -0.021, with correlation coefficient R2 = 0.951 and 0.994, respectively. The correlation between measured taste score, moisture content and protein content of AN800 machine of milled rice was TAT=127.352+1.550TAM -13.603TAP, with correlation coefficient R2=0.739. The relationship of taste scores in overall flavor between the panel tests and the AN800 machine was TT=-0.070TA3-0.408TA2+0.529TA+0.014, with correlation coefficient R2=0.803. The developed calibration curves from rice NIR reflection spectrum to measure moisture (10-15%, w.b.) and protein contents (4.9-6.6%, d.b.) of brown and milled rice can be further applied to develop NIR sensory taste instrument.
摘 要 稻米含水率及蛋白質含量為影響其食味品質的重要因素。本研究使用近紅外線反射光譜檢測稻米含水率及蛋白質含量之相關變化,並配合食味檢測建立檢量線,以作為非破壞性快速檢測之依據。稻米含水率校正是以市售之台稉9號稻米調製成不同含水率(10-16%, w.b.)之樣本測試;稻米蛋白質含量是以台稉9號及台農71號之稻米樣本在生長後期施加不同氮肥 (N=0、40、80、120、160與200 kg/ha)以改變其蛋白質含量。 以烤箱法驗證AN800量測稻米(台稉9號)含水率成分之準確性,兩者之關係式:糙米含水率mAB=0.878mOB+2.243及白米含水率mAM=1.109mOM-0.117,其判別係數R2為0.987及0.979。利用定氮儀(Kjetec 2400)驗證AN800量測稻米蛋白質含量之準確性,兩者之關係式:糙米(台農71號)蛋白質含量pAB=0.775pKB+1.516及白米(台稉9號)蛋白質含量pAM=1.027pOM-0.289,其判別係數R2為0.916及0.979。 台稉9號稻米(糙米、白米)之近紅外線光譜波長範圍700-2500 nm以MPLS回歸分析與AN800檢測相關之檢量線分析顯示,與含水率相關檢量線之R2為(0.997、0.999)、與蛋白質含量相關檢量線之R2為(0.955、0.967)及與食味值相關檢量線之R2為(0.990、0.969),其對應之校正誤差SEC分別為(0.090、0.056)、(0.094、0.109) 及(0.655、1.148),對應之預測誤差SEP分別為(0.111、0.088)、(0.148、0.162)及(1.554、1.270)。 台稉9號官能品評結果顯示,其白米含水率與所煮米飯食味品評之外觀、黏性和總評分數成正相關,與硬性分數成負相關。白米蛋白質含量與米飯之官能品評之口味、黏性和總評分數成負相關,與硬性分數成正相關。米飯官能品評總評分數相關於AN800量測之白米食味值,其含水率(10-15%, w.b.) TMT=1.292TMA3+1.615TMA2+ 1.341TMA-0.068,蛋白質含量TPT=0.330TPA3-0.027TPA2+0.206TPA -0.021,判別係數R2分別為0.951及0.994。AN800所檢測白米食味值與含水率及蛋白質含量(4.9-6.6%, d.b.)之函數關係式為TAT=1.550TAM -13.603TAp+127.352,判別係數R2=0.739。官能品評之總評分數相關於AN800量測之食味值,其關係式為 ,判別係數R2=0.803。 由稻米NIR反射光譜所得之含水率(10-15%, w.b.)及蛋白質含量(4.9-6.6%, d.b.)檢量線可以快速預測糙米及白米含水率及蛋白質含量,此研究結果可作為發展NIR食味檢測儀之依據。
Appears in Collections:生物產業機電工程學系



Show full item record
TAIR Related Article

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.