查詢結果分析
來源資料
頁籤選單縮合
題 名 | 洋香瓜糖度檢測之研究--(2)近紅外線分析法=Determination of Sugar Content in Cantaloupe--(Ⅱ)Near Infrared Spectroscopy |
---|---|
作 者 | 張文宏; 陳世銘; 郭立穎; | 書刊名 | 農業機械學刊 |
卷 期 | 7:1 1998.03[民87.03] |
頁 次 | 頁87-98 |
分類號 | 434.251 |
關鍵詞 | 糖度; 近紅外線; 洋香瓜; 果汁; 果肉; Sugar content; Near infrared spectroscopy; Cantaloupe; Juice; Flesh; |
語 文 | 中文(Chinese) |
中文摘要 | 本研究利用近紅外線光譜檢測洋香瓜果汁及果肉之糖度,探討之波長範圍為1000nm 至2500nm,而糖度範圍為5.5至13.2°Brix。在果汁透射光譜分析中,2274nm,2241nm,2321nm 三個波長不論是在原始光譜(log(1/T))或二次微分光譜(D□log(1/T))中,均對洋香瓜之糖度預 測有極大的影響;由二次微分光譜2274nm,U43nm,224lnm以及232lnm四個波長組合之校正 方程式對預測樣本之相關係數可達0.970,而標準預測誤差(SEP)及相對預測誤差(RSEP)為 0.383及3.64%。果肉反射光譜之原始光譜對於糖度的預測結果並不理想,但是其二次微分 光譜(D□log(1/R))與其糖度值則有極佳的線性關係;試驗結果顯示,由1688nm, 232lnm,140lnm,2000nm及1106nm五個波長所組成的多重線性迴歸校正方程式,其標準校正 誤差(SEC)為0.728,相關係數為0.917,對於預測樣本的預測,其標準預測誤差(SEP)及相對 預測誤差(RSEP)分別為0.769及7.21%,相關係數則為0.892。不論果汁或果肉之糖度檢測, 二次微分光譜顯然較原始光譜為佳,尤其以單波長校正方程式更為明顯。 |
英文摘要 | Near infrared spectroscopy was used to determine the sugar content in the juice and flesh of cantaloupe. The investigation was conducted in the wavelength region from 1000 to 2500 nm, and the sugar content was ranged from 5.5 to 13.2° Brix. The analysis of juice transmittance spectra showed that wavelengths 2274,224l and 232l nm were very effective in predicting sugar content of cantaloupe juice in both original absorbance and the second derivative spectra. The juice calibration equation with 4 wavelengths (2274,1143,224l and 232l nm) could predict the sugar content of samples in the prediction group with a linear correlation coefficient of 0.970, a standard error of prediction (SEP) of 0.383 and a relative standard error of prediction (RSEP) of 3.64%. Regarding the reflectance spectra of cantaloupe flesh, the ability of predicting sugar content by the second derivative spectra was much better than that of original spectra. The calibration equation developed by multiple linear regression, which consisted of 5 wavelengths (i.e. 1688,2321,1401,2000 and 1106 nm) gave a correlation coefficient of 0.917 and standard error of calibration (SEC)of 0.728 for the calibration group, while a correlation coefficient of 0.892,an SEP of 0.769 and an RSEP of 7.2I% were reported for the prediction group. For the determination of sugar content in both juice and flesh, the calibration equations developed by using the second derivative spectra were better than those by using original spectra, it was especially obvious for those calibration equations which consisted of only single wavelength. |
本系統中英文摘要資訊取自各篇刊載內容。