A method and a device for generating a three-dimensional virtual environment, wherein the method includes: acquiring a plane environment image captured by an image capturing device; extracting geometric information from the plane environment image to obtain geometric information of each two-dimensional element in the plane environment image; extracting content information from the plane environment image to obtain content information of each two-dimensional element in the plane environment image; respectively determining three-dimensional models corresponding to the two-dimensional elements according to the geometric information and the content information of each two-dimensional element; respectively rendering three-dimensional models corresponding to the two-dimensional elements at coordinates corresponding to the geometrical shape of each two-dimensional element and generating a three-dimensional virtual environment.