In order to solve the problems of low accuracy,high computational cos t and failure to make full use of space spectrum information of traditional hyperspectral image classi fication methods,a hyperspectral image classification method based on multi-dimensional parallel c onvolution neural network (3D-2D-1D PCNN) is proposed in this paper.Firstly,the algorit hm uses different dimensions of convolutional neural network (CNN) to extract the spatial spectral features,spatial features and spectral features of hyperspectral image information.Then,the same parallel convolution layer is us ed to fuse the combined spatial spectral features,spatial features and spectral features.Finally,hype rspectral image information is accurately classified by linear classifier.The proposed method can not only extract the deeper spatial and spectral feature information in hyperspectral images,but also fuse the feat ures of different dimensions of spectral images to reduce the computational cost.Comparative expe riments are carried out on Indian Pines,Pavia Center and Pavia University data sets.The results sh ow that the proposed algorithm obtains the optimal results,and the classification accuracy reaches 9 9.210%,99.755% and 99.770% respectively.