Abstract: To address the issue that existing LiDAR-based material flow detection methods for belt conveyors are susceptible to abnormal point cloud data and struggle to accurately describe the surface state of materials, a laser-based material flow detection method for belt conveyors using Akima interpolation is proposed. The method involved acquiring point cloud trajectories of the conveyor belt using LiDAR, followed by pass-through filtering and outlier noise removal. The Akima interpolation method was then used to obtain the cross-sectional area of material on the belt. Combined with the conveyor's operating speed and LiDAR scanning frequency, the material volume within a single scan cycle was calculated. By integrating the measurement data over any given time period, the total material volume during that period could be obtained. Simulation results showed that denoising outlier points from the LiDAR output point cloud could effectively identify and correct abnormal data, resulting in calculated values that were closer to the actual material cross-sectional area. Comparative experiments using both the sector-triangle calculation method and the Akima interpolation method under varying volumes and belt speeds demonstrated that the sector-triangle method had lower and less stable accuracy, while the Akima interpolation method consistently achieved accuracy above 90%, offering high reliability and enabling accurate measurement of both instantaneous and total material flow.