In order to study the classification of ancient porcelains based on the physical structures of glaze layer, optical coherence tomography (OCT) technology combined with X-ray fluorescence spectroscopy (XRF) technology, scanning electron microscope-energy dispersive spectroscopy (SEM-EDS) technology and laser Raman spectroscopy (LRS) technology are used to analyze the Celadon and Jun porcelains which are dated to the Jin and Yuan dynasties and excavated from the Qingliangsi site in Baofeng County, Henan Province, China. The glaze layers are classified qualitatively according to the characteristics of OCT images of glaze layers. The image texture analysis technique is used to characterize the OCT images of glaze layers quantitatively, and principal component analysis is performed based on the image texture parameters determined by the image texture analysis. The classification results of the enamel according to the OCT image are compared with the classification results of the chemical composition of the glaze obtained from XRF. The relationship between the chemical compositions of glaze layers and the OCT image feature are discussed based on the results of SEM-EDS and LRS.