Energy efficiency can be obtained by measuring accurate new surface energy in the crushing process. For the calculation of new surface energy, most researchers only pay attention to particle size distribution and ignore the influence of particle shape and roughness on the surface area. In this paper, the image processing technology was used to calculate the shape parameters and surface fractal dimension of the crushed granite sample. According to the different combinations of particle shape and roughness, the new surface area corresponding to the four basic models was calculated. For the surface area of a single particle, the calculation result of the rough model considering the surface fractal dimension is higher than that of the smooth model. Moreover, the ratio of the calculation results of the rough model and the smooth model increases significantly as the particle size increases. For 0.1 mm particles, the area ratio of the two models is 8, but for 25 mm particles, the area ratio reaches 130. In contrast, the particle shape is a secondary factor that affects the surface area calculation. The ellipsoidal model considering the particle shape has a surface area 30% larger than the spherical model. If the roughness and particle shape are considered when calculating the surface area, the energy efficiency of crushing is higher.