3-D object recognition using a genetic algorithm

The multiple-view approach is the most typical approach for the 3-D object recognition. This approach models each 3-D object by a collection of 2-D projections from various viewing angles. An advantage of the multiple-view approach over other approaches for 3-D object recognition is that it can be applied for the recognition of 3-D objects which have relatively complex shape. However, if the target objects are complex in shape, the size of the model database becomes large. As a result, the approach requires long time for the recognition of such objects. In this paper we present a 3-D object recognition algorithm based on the multiple-view approach which can find a best matched model by only searching a small percentage of the model database. The algorithm uses a genetic strategy to select a best matched model from the database.