论文部分内容阅读
目的通过近红外光谱技术对苹果进行无损检测,具有快速、非破坏性、无试剂分析、安全、高效、低成本等特点。方法本文运用Matlab7.1,对苹果的糖度与光谱曲线在不同波段处的相对光强进行多元线性回归分析,建立了最佳单波长、最佳双波长组合、最佳三波长组合和最佳四波长组合的校正方程。结果校正方程的相关系数分别为0.621、0.715、0.797和0.822,标准校正误差分别为1.325、1.218、1.074和0.997Brix°,相对校正误差分别为11.28%、10.42%、9.63%和9.26%。结论试验表明用近红外光谱技术无损检测苹果糖度是可行性的,为今后进一步检测苹果的其他品质奠定基础。
Aim To study the nondestructive testing of apple by near infrared spectroscopy, which has the characteristics of fast, nondestructive, reagentless analysis, safety, high efficiency and low cost. Methods Using Matlab7.1, multiple linear regression analysis was performed on the relative intensities of apple brix and spectral curves in different wavelength bands. The optimal single wavelength, the best dual wavelength combination, the best three wavelength combination and the best four Correction equations for wavelength combinations. Results Correlation coefficients of calibration equation were 0.621, 0.715, 0.777 and 0.822 respectively. The standard calibration errors were 1.325, 1.218, 1.074 and 0.997Brix °, respectively. The relative correction errors were 11.28%, 10.42%, 9.63% and 9.26% respectively. Conclusion Experiments show that it is feasible to use non-destructive detection of apple sugar content by near-infrared spectroscopy, which lays the foundation for further testing of other apple quality in the future.