The Raman intensity of different biological samples or slices at the same wavenumber would be probably different when repeatedly detected at the same location,which was due to fine variation of focused position and the system deviation and made quantitative analysis much hard.Consequently,it was necessary to get a more feasible approach of defining internal or external standard for quantitative detection,which would be helpful for comparing Raman spectra captured under different conditions directly.Under the condition of polarization and depolarization,the Raman spectra of canine hip articular cartilage slices were detected and processed by standard normal variate (SNV) and multi-scattering correction (MSC) methods.In addition,eosin used as a novel internal standard to preprocess Raman signal was discussed and explored.It was found that the MSC method was most suited for the depolarization condition,as well as 1501 cm-1 band of eosin as internal standard for the polarization condition.The Raman band of eosin didn′t show anisotropy with polarization angel,which was effective to use as internal standard for the normalization of Raman spectra of biological tissue.Further,the internal standard method based on eosin brought more simple operation and more accurate results than MSC and SNV.Therefore,this study provided a promising novel method for the quantitative analysis of biological samples based on Raman spectroscopy.