为了寻求一种快速、无损检测“深州蜜桃”果实内部品质的方法，该研究中采用SACMI近红外分析仪以漫反射方式对采收期果实可溶性固形物含量（Soluble Solid Content，SSC）、硬度、pH进行了无损检测，并与传统参考方法所测结果进行比较。结果表明，剔除异常值后，SSC无损检测预测值与传统参考方法实测值的两组数据相关性（R2=0.79，p<0.01，SEP=0.47）优于硬度（R2=0.47，p<0.01，SEP=2.01）及pH（R2=0.40，p<0.01，SEP=0.14）；SSC模型（RPD=2.15，RMSEP=0.79%，RSD=6.2%）预测性能较高，可满足于快速检测，pH模型（RPD=1.29，RMSEP=0.16，RSD=3.1%）、硬度模型（RPD=1.37，RMSEP=2.37 kg/cm2，RSD=39.4%）预测性能较低。研究表明：采用近红外分析仪检测“深州蜜桃”果实SSC是可行的，可为果实内部品质的评价提供了实时、快速、无损的检测方法，进而为其智能分级提供理论基础与技术支持。

To develop a rapid and non-destructive method for detecting the internal quality of the “Shenzhou Peaches”, the SACMI near-infrared analyzer was used in the diffuse reflection for non-destructive measurements of the soluble solid content (SSC), firmness, and pH of fruits ready for harvest. The results were compared with those obtained by traditional methods. The results reveal that, after excluding outliers, the correlations between the predicted SSC values by the non-destructive method and the measured values using traditional methods (R2=0.79, p<0.01, SEP=0.47) are better than those of firmness (R2=0.47, p<0.01, SEP=2.01) and pH (R2=0.40, p<0.01, SEP=0.14). The SSC model (RPD=2.15, RMSEP=0.79%, RSD=6.2%) exhibits better prediction performances and can enable rapid detection. Meanwhile, the pH model (RPD=1.29, RMSEP=0.16, RSD=3.1%) and the firmness model (RPD=1.37, RMSEP=2.37 kg/cm2, RSD=39.4%) gave less satisfactory prediction results. This research has shown that detecting the SSC of “Shenzhou Peach” fruits is feasible with a near-infrared analyzer. This is a real-time, fast and non-destructive method for the internal quality evaluation of fruits. This finding can provide a theoretical foundation and technical support for intelligent future grading and classification of fruits.

河北省农林科学院创新工程项目（2019-2-1-1）；河北省现代农业产业技术体系项目（HBCT2018100207）