Abstract
We have been proposed the algorithm for the object recognition of individual objects in a recycling system. When this method is extended to the recognition of multiple objects, it is possible to place the reference images in a data base and try to match each objects to the reference images in the data base. However, as the number of reference images increases, the process of matching an object to every reference images becomes tedious. In this paper, we propose the recognition algorithm using an associative network so that the search for the most probable matching pattern is efficiently carried out.
