In order to further improve the classification performance of sparse representation classification, a hyperspectral image (HSI) classification algorithm based on joint sparse representation with morphological feature extraction is proposed. To obtain the principle component images, the whole HSI is analyzed by principle component analysis. The closing and opening operations are implemented on principle component images to extract the morphological features. Combining the original spectral and the morphological feature, the pixels in a local region around the central test pixel are simultaneously represented by a set of common atoms of new training dictionary. The classification of HSI is determined by computing the minimum reconstruction error between testing samples and training samples. Experimental results on AVIRIS and ROSIS HSI demonstrate that the effectiveness of the proposed method for improving the classification accuracy and performance.