本文采用不同方法对来自正常和病变关节软骨样本的红外光谱进行预处理，而后利用主成分分析对关节软骨进行鉴别分析。首先对关节软骨切片实现傅里叶变换红外光谱采集，其次分别采用基线校准、标准化、多元散射校正和标准正态变量变换对软骨的红外光谱进行预处理，然后对原始光谱（矩阵）以及预处理光谱进行主成分分析，根据得分矩阵对样本进行分类分析。结果表明：预处理方法结合主成分分析可以更好地对正常和病变关节软骨样本进行分类，而且多种预处理方法的结合可以更好地增强样本间的区分度。另外，针对关节软骨样本，多元散射校正比标准正态变量变换具有更好的增强效果。

In this paper, different methods were used to preprocess the infrared spectra of healthy and osteoarthritic articular cartilage. And then the principal component analysis (PCA) was applied to classify the articular cartilage samples. First, FTIRI on articular cartilage specimens was achieved. After extracted from the FTIR images, the infrared spectra were preprocessed by baseline correction, normalization, multiplicative scatter correction and standardized normal variate. And then, the original and preprocessed spectra were imported into SPSS software for PCA and the cartilage samples were classified based on the score marix. It is indicated that preprocessing methods combined with the principal component analysis can classify the samples of healthy and osteoarthritic articular cartilage, better. And the result of the combination of different methods is better. In addition, for the cartilage samples, the influence of multiple scatter correction is better than standardized normal variate.