Curve and surface reconstruction based on unorganized data points plays an important role in the fields of reverse engineering and computer vision. This paper presents a new algorithm based on SOM to realize Cubic B-spline reconstruction. A set of unorganized data points and an initial neural network are given. The network can be optimized using the algorithm to make the neurons gradually approach the given unorganized data points and reflects spatial characteristics of unorganized data points. Control poin...