In response to the deviation between the 3D design model and the actual situation, this article studies the difference and color texture correction techniques between point clouds and design models, as well as the updating and correction methods of 3D design models based on point cloud models. Obtain the point cloud model corresponding to the design 3D model based on the topological relationship of the 3D design model, and use steps such as initial registration, accurate registration, and color labeling to match the overlapping parts of the two model point clouds through geometric registration. Record the points in the non-overlapping areas as difference points, and display the geometric surface where the difference points are located as difference points. This paper studies the parameterized correction of equipment model points under the constraint of point clouds, and studies the method of correcting the position, posture, and size of power tower models under the constraint of completion acceptance or scanning point clouds by transportation and inspection departments. The proposed improved ICP algorithm based on three-dimensional topological relationships overcomes the disadvantage of the original algorithm requiring initial registration values in advance, and is suitable for point cloud scenes that have already been classified, with good practical value. Parametric characterization is performed on all iron towers on the entire transmission line to obtain information about the type, tower type, insulator, point location, and elevation of each tower. Aiming at the problem that it is difficult to update and correct the overall 3D model at one time, the overall design model is stored and organized as a unit of equipment, and the tower correction method is iteratively updated. Finally, all line tower correction results are obtained, achieving the revision of the engineering level 3D model layout structure.