In order to overcome the inefficiency of traditional stockpile volume measurement methods, a 3D laser scanning technology based stockpile modeling system was designed and implemented, which allows the operator to quickly make a characteristically-detailed 3D model by circling the stockpile with the scanning device in hand. The scanning device uses a 2D laser scanner to scan the stockpile with high-precision sensors to track its position and orientation, and then integrates discrete scanning points into a 3D point cloud for modeling. The algorithm optimization was conducted upon the characteristics of the scanner and the stockpile to reduce the time consumed for the noise reduction of scanning points and to improve the efficiency of 3D modeling and volume calculation. As shown by the measuring experiment of several 3000～30000 m3 coal piles, the 3D modeling of the stockpile can be completed in a few minutes which mainly dependent on the walking speed of the operator circling the stockpile. The 3D model and its volume can be acquired within 20 seconds after the measurement completes, and the accuracy is within 0.75%.