Landmark-based shape recognition by a modified Hopfield neural network

A new method is introduced to achieve partial shape recognition by means of a modified Hopfield neural network. Given a scene consisting of partially occluded objects, a model object in the scene is hypothesized by matching the landmarks of the model with those in the scene. A local shape measure, known as the sphericity of a triangular transformation, is used as a measure of similarity between two landmarks. The hypothesis of a model object in a scene is completed by matching the model landmarks with the scene landmarks. The landmark matching task is performed by a modified Hopfield neural network. The location of the model in the scene is estimated with a least squares fit among the matched landmarks. A heuristic measure is then computed to decide if the model is in the scene