An image generating method, includes: establishing a 3D scene model that includes a virtual 3D object, a virtual display screen and at least one sight point set; determining a plurality of intersection points between a plurality of virtual light paths from each sight point to a plurality of virtual object points on the surface of the virtual three dimensional object and the virtual display screen, all virtual light paths {lk1, lk2, . . . , lkS} corresponding to the virtual object point Tk intersect at the virtual object point Tk; wherein, 1≤k≤n, 1≤i≤S, S is the total number of established sight points, and the plurality of virtual object points are all located within the viewing angle range of the virtual display screen; forming S frames of rendered images, including: determining color parameters of a plurality of intersection points on the virtual display screen to obtain the ith rendered image.