The invention relates to the field of point cloud data processing. A main traveling direction between a start point ps(xs,ys,zs) and an end point pe (xe,ye,ze) of the geodesic path is selected, and agridding area is determined by the main traveling direction; coordinate values of all data points in the gridding area are extracted, three coordinate component arrays are formed, the arrays are sorted, the same coordinate component values are removed, and maximum and minimum values of the coordinates of all point cloud data points in the gridding area are found; step sizes hxi, hyj and hzk of thedata points in three axial directions are calculated respectively, the data points between the two points ps and pe are subjected to non-uniformly gridding, and the step 3 is executed; arrival time of each cell in the wavefront grid is calculated by UNCDFMM according to the step sizes hxi, hyj and hzk; the cells meeting the approximate geodesic property are screened out after the arrival time ofeach cell is calculated, and vertices of the cells are sequentially connected to form the geodesic path.