中文摘要：

西瓜是一种广受世界各国消费者喜爱的水果，坚实度是西瓜的一个重要品质指标，文章利用可见／近红外漫透射光谱技术进行了西瓜坚实度（FM）的无损检测研究。采用偏最小二乘法（PLS）和主成分回归法（PCR）建立了FM与漫透射光谱的无损检测数学模型，对比分析了不同光谱预处理方法（原始光谱％T，-阶微分处理光谱D^1（％T），二阶微分处理光谱D2（％D以及光谱的Savitsky-Golay法滤波）对模型预测性能的影响。根据模型相关系数（r）及预测平方根标准偏差（RMSEP）进行了不同模型的预测性能对比，结果表明：光谱经二阶微分处理并使用Savitsky-Golay法滤波后，采用PLS法可以得到最好的FM建模结果（r=0．974，RMSEP=0．589N）。研究表明：应用可见／近红外漫透射光谱技术检测西瓜的坚实度是可行的，为今后快速无损评价大果形厚果皮类水果坚实度提供了理论依据。

英文摘要：

Watermelon is a popular fruit in the world and firmness （FM） is one of the major characteristics used for assessing watermelon quality. The objective of the present research was to study the potential of visible/near Infrared （Vis/NIR） diffuse transmittance spectroscopy as a way for the nondestructive measurement of FM of watermelon. Statistical models between the spectra and FM were developed using partial least square （PLS） and principle component regression （PCR） methods. Performance of different models was assessed in terms of correlation coefficients （r） of validation set of samples and root mean square errors of prediction （RMSEP）. Models for three kinds of mathematical treatments of spectra （original, first derivative and second derivative） were established. Savitsky-Goaly filter smoothing method was used for spectra data smoothing. The PLS model of the second derivative spectra gave the best prediction of FM, with a correlation coefficient （r） of 0. 974 and root mean square errors of prediction （RMSEP） of 0. 589 N using Savitsky-Goaly filter smoothing method. The results of this study indicate that NIR diffuse transmittance spectroscopy can be used to predict the FM of watermelon. The Vis/NIR diffuse transmittance technique will be valuable for the nandestructive detection large shape and thick peel fruits＇.