Rapid emergency rescue measures are essential in reducing losses resulting from disasters. With the development of science and technology, remote sensing technology is gradually applied to emergency rescue work, especially SAR remote sensing technology, with its advantages of all-weather and all-time observation, which play an important role in emergency rescue. In this paper, we make use of the Polarization information and texture information of the full Polarimetric SAR(PolSAR) remote sensing image after the Yushu earthquake to accurately extract information of intact and collapsed buildings. Firstly, we separately extract the ridge and hillside information form (1-H)(1-A) and Yamaguchi Decomposition by method of CNN (Convolution Neural Network). Secondly, we successfully extract building area (Intact Building Area +Non-Building Area) combine Homogeneity and Second Moment based on previous information of ridge and hillside. Finally, extract the intact building and collapsed building from building area based on Anisotropy and Correlation by method of CNN (Convolution Neural Network) again. The accuracy of extractive information is 99.12% by method of CNN.