Pre-processing is essential for processing the row data point clouds which acquired using a 3D laser scanner as a modern technique to digitize and reconstruct the surface of the 3D objects in reverse engineering applications. Due to the accuracy limitation of some 3D scanners and the environmental noise factors such as illumination and reflection, there are some noised data points associated with the row point clouds, so, in the present paper, a preprocessing algorithm has been proposed to determine and delete the unnecessary data as noised points and save the remaining data points for the surface reconstruction of 3D objects from its point clouds which acquired using the 3D laser scanner (Matter and Form). The proposed algorithm based on the assessment of tangent continuity as a geometrical feature and criteria for the contiguous points. A MATLAB software has been used to construct a program for the proposed point clouds pre-processing algorithm, the validity of the constructed program has been proved using geometrical case studies with different shapes. The application results of the proposed tangent algorithm and surface fitting process for the suggested case studies were proved the validity of the proposed algorithm for simplification of the point clouds, where the percent of noised data which removed according to the proposed tangent continuity algorithm which achieved a reduction of the total points to a percentage of (43.63%), and (32.01%) for the studied case studies, from the total number of data points in point cloud for first and second case study respectively.