In order to measure the stacked volume of filamentary material on the conveyor belt in real-time and accurately, a laser sensing technology was used to establish an online measurement method, combined with the smoothing algorithm of material surface contour point cloud data. Theoretical analysis and experimental verification were conducted to obtain the measurement accuracy and repeatability data of stacked volume after processed by three smoothing algorithms. The results show that, the height, width and cross-sectional area of the material accurately can be obtained, but the repetitiveness of volume is bad, whose variation coefficient is close to 5%. The smoothing effect of Moving algorithm is best. After the contour data is processed by Moving algorithm, the range and variation coefficient of volume decreased to 105.20 cm³ and 1.36% respectively, the single accuracy and multiple repetitiveness are both greatly improved. The results are helpful to improve the control stability and various indicators of process in cigarette production.