The surface fitting method of complex surface object was researched, which inclued the processes of parameterization of point cloud data, immune genetic algorithm (IGA) based adaptive node calculation, reversing of curve control points，segmentation of point cloud and surface stitching constitute surface fitting. Fusion of immune genetic algorithm and ant colony system fitting algorithm was proposed in surface. The immune genetic algorithm was introduced into the iterative process of each ant colony, and the ability of local optimization and global search ability of immune genetic algorithm can improve the convergence speed of ant colony. In the algorithm, the immune genetic algorithm can effectively overcome the shortcomings of ant colony system which easily trap into local optimal solution or degradation by selecting, intersecting and mutating of genetic algorithm along with the immune vaccination of immune algorithm. The fender，fandisk, bunny and cow of the 3D scattered points cloud were fitted by fusion algorithm, genetic algorithm (GA) and immune genetic algorithm (IGA). The experiments showed that the new algorithm was very excellent in convergence speed and global optimal solution search capability. The fitting surface of this algorithm was better than those of GA and IGA in fitting precision with fitting precision improved by 18% and 11%, respectively, which can meet the requirements of complex curved surfaces.