Shape recognition by artificial neural network for validating coins.

摘要

A SYSTEM OF CURRENCY VALIDATION TO DETERMINE IF THE CURRENCY THAT MOVES ALONG A RAIL IS VALID CURRENCIES AND IF IT IS, YOUR NAME. The system includes a RAIL TO THE LAKE IS currencies move, at least one photodetector placed at ALONG THE RAIL TO DETECT THE PRESENCE OF ABSENCE OF A COIN MOVING therethrough, at least one magnetic sensor associated with each DETECTOR OPTICAL PLACED NEAR THE DETECTOR OPTICAL correspondingly, WHERE EACH auto switches includes an inductive element and a circuit for driving the detector MAGNETICO to generate a field that is coupled to the coin passing so that CURRENCY AND THE ELEMENT INDUCTION HAVE A mutual inductance between, WHERE THE CIRCUIT PUTS RING TO DETECTOR MAGNETIC A predetermined number of times WHILE THE cURRENCY IS NEXT TO THE MAGNETIC DETECTOR WITH WHAT THE DETECTOR MAGNETIC CREATES A damped wave which has a representative characteristics of PHYSICAL AND MAGNETIC COIN, a preprocessor SIGNAL CHARACTERISTICS LINKED TO DETECTOR MAGNETIC operably PA RA GENERATE output responses representative of distinctive features of the coin circuit feature extraction TO EXTRACT OF RESPONSES preprocessor output SIGNAL PORTION OF SIGNALS REPRESENTATIVE OF FEATURES DISTINCTIVE DEFAULT COIN, a circuit for generating a representation MULTIDIMENSIONAL CHARACTERISTICS LEARNED AND TO COMPARE THE REPRESENTATION MULTIDIMENSIONAL WITH THE CENTER OF A GROUP ELLIPSOIDAL STATED coin denominations SELECTED TO DETERMINE THE SCOPE OF THE COMPARISON BETWEEN THEM AND BE USED TO DETERMINE IF THE CURRENCY IS AN ACCEPTABLE CURRENCY OR NOT, AND CIRCUIT SORT OF ARTIFICIAL NEURAL NETWORK CONNECTIONS HAVING SOME preprocessor AND COMPARISON CIRCUIT, WHERE THE CIRCUIT ARTIFICIALTIENE NEURAL NETWORK SORT OUT AN IDENTIFYING THE NAMING OF COINS cOMPARATOR CIRCUIT HAS DETERMINED THAT ARE ACCEPTABLE.