| 摘要 |
PROBLEM TO BE SOLVED: To automatically select shoes fitted to feet in place of techniques or experiences of shoe fitter by efficiently using data automatically measured by a machine. SOLUTION: The shape of foot is measured by a gauge and the feature data of two-dimensional shape on the front, lateral and up sides of respective toe parts, central parts, heel parts are inputted to respective neural networks. When the upper feature data of the toe parts of shoe die are found from the upper feature data of toe parts by the neural network, for example, the upper model of toe part of the shoemakers last suitable for these feature data is selected, similarly, the side and front models of toe parts are selected, and the models of toe parts of the shoe last matched with these upper, side and front models of toe parts of the shoe last are found. Similarly, the models of central part and heel part of the shoe last are found, and the code of the shoe last coincident with these models of toe part, central part and heel part of the shoe last is selected from these models and outputted to a display device. |
