METHOD AND SYSTEM FOR MATERIAL CLASSIFICATION OF AN OBJECT

    公开(公告)号:US20230314317A1

    公开(公告)日:2023-10-05

    申请号:US17709679

    申请日:2022-03-31

    CPC classification number: G01N21/359 G01N2201/126

    Abstract: A method for material classification of an object. The method contains the steps of measuring, by a first light sensor, an intensity of a first light signal from an object; measuring, by a second light sensor, an intensity of a second light signal from the object; inputting features related to the intensity of the first light signal and features related to the intensity of the second light signal into a neural network; calculating by the neural network, for each of a plurality of possible classes of material, a probability of the object containing the material of the class; and determining a material of the object based on one of the plurality of possible classes of material that has a highest probability. Embodiments of the invention provide an automatic solution of material classification based on neural network, which achieves a high successful recognition rate of the material.

Patent Agency Ranking