Abstract: To address the issues of low automation, strong manual intervention, and high data dispersion in traditional roadway surrounding rock deformation monitoring methods, a deformation monitoring approach for roadway surrouding rock based on 3D point cloud data acquisition technology is proposed in this paper. This method begins with noise reduction through a combined "statistical-median filtering" method during point cloud preprocessing. Then a voxel grid downsampling method integrating density constraints and centroid neighboring points is introduced to streamline the point cloud data. Subsequently, the roadway centerline is calculated using the bidirectional projection method and point cloud processing software is employed to select point cloud data, enabling the acquisition of roadway cross-sections at equal intervals and thicknesses and fitting their contour curves. Finally, a combined holistic and local monitoring analysis approach is employed to assess roadway surrounding rock deformation. Verified through field engineering applications, this method can adapt to complex surface conditions of roadway surrounding rock and realize millimeter-level measurement precision for surface displacement. It can effectively evaluate the rock deformation of the test working face and fully meet the requirements for surface deformation monitoring of surrounding rock in mine roadways.