In order to characterize the three-dimensional texture of coarse aggregates, the Laplacian, distance-weighted Laplacian, cotangent-weighted Laplacian and Taubin smoothing algorithms are considered in this study. Two smoothing scale parameters are added in the algorithms to control the smoothing operation within the texture level. Two standard geometric models (sphere and cube) are constructed to compare the effectiveness of the four smoothing algorithms in removing the surface texture. The results show that the Taubin smoothing algorithm has an obvious advantage over the other three algorithms. The three parameters (smoothness distance, smoothness level and negative scaling factor) in the Taubin smoothing algorithm are further investigated. Based on the study, the smoothness distance is suggested to be between one time the size of mesh and 2 times the size of mesh and should be determined according to the parameters analysis results. The smoothness level is suggested to be 1.1 and the negative scaling factor is suggested to be −1.5. The advantage of the two smoothing scale parameters is also studied. The results show that the two smoothing scale parameters can effectively control the smoothing operation within the given texture level. Finally, the effectiveness of the Taubin smoothing algorithm in coarse aggregate texture characterization is verified by the actual coarse aggregate models. The results demonstrate that the Taubin smoothing algorithm can effectively distinguish the texture of coarse aggregates and can be used to characterize the coarse aggregate texture.