Inspection device
An inspection device inspects of goods that include plural articles that sometimes overlap each other. An X-ray inspection device irradiates goods containing plural articles having a predetermined shape and inspects the goods on the basis of inspection images obtained from radiation that has passed through the goods or radiation that has reflected off the goods. The X-ray inspection device includes a storage component, a learning component, and an inspection component. The storage component stores, as teaching images, at least the inspection images of the goods that are in a state in which the plural articles overlap each other. The learning component acquires, by machine learning using the teaching images stored in the storage component, features relating to the goods that are in a state in which the plural articles overlap each other. The inspection component inspects the goods using the features that the learning component has acquired.
1. An inspection device that irradiates, with radiation, goods containing plural articles having a predetermined shape and inspects the goods on the basis of inspection images obtained from radiation that has passed through the goods or radiation that has reflected off the goods, the inspection device comprising:
an electronic controller having a CPU and electronic memory that includes a storage component that stores teaching data that correlates the inspection images of the goods that are in a state in which the plural articles overlap each other with overlapping region data and contaminated region data of the inspection images;
the electronic controller further includes a learning component that acquires, by machine learning using the teaching data stored in the storage component, features of the inspection images extracted from the teaching data, as features relating to the goods that are in a state in which the plural articles overlap each other, the features of the inspection images being the features of the inspection images of the goods that include contaminants and the features of the inspection images of the goods that do not include contaminants; and
the CPU is connected to an inspection irradiator and sensors that inspects for the presence or absence of contaminants included in the goods using the features that the learning component has acquired,
the overlapping region data are data relating to regions where the plural articles overlap each other in the inspection images,
the contaminated region data are data relating to regions where there are contaminants in the inspection images.
2. The inspection device according to claim 1 , wherein
the storage component stores, as the teaching images, at least the inspection images of the goods that include contaminants, and
the learning component acquires the features by machine learning using the teaching images and regions in the teaching images where there are contaminants included in the goods.
3. The inspection device according to claim 1 , wherein
the storage component stores, as the teaching images, at least images in which images of virtual contaminants have been incorporated into the inspection images of the goods that do not include contaminants, and
the learning component acquires the features by machine learning using the teaching images and regions in the teaching images where there are the virtual contaminants.
4. The inspection device according to claim 1 , wherein the inspection irradiator further inspects the quantities of the articles contained in the goods.
5. The inspection device according to claim 4 , wherein the learning component acquires the features by machine learning using the teaching images and the quantities of the articles contained in the goods in the teaching images.
6. The inspection device according to claim 1 , wherein the inspection irradiator acquires a final inspection result of the goods on the basis of a plurality of inspection results comprising first inspection results of the goods in a case where it has used the features acquired by the learning machine and second inspection results of the goods in a case where it has not used the features acquired by the learning machine.
7. An inspection device that irradiates, with radiation, goods containing plural articles having a predetermined shape and inspects the goods on the basis of inspection images obtained from radiation that has passed through the goods or radiation that has reflected off the goods, the inspection device comprising an electronic controller having a CPU and electronic memory:
the electronic memory includes a storage component that stores teaching data that correlates the inspection images of the goods that are in a state in which the plural articles overlap each other, with overlapping region data and contaminated region data of the inspection images;
the electronic controller includes a learning component that acquires features of the inspection images extracted from the teaching data, as features relating to the goods, the features of the inspection images being features of the inspection images of the goods that include contaminants and features of the inspection images of the goods that do not include contaminants; and
the CPU is connected to an inspection irradiator and sensors that inspects for the presence or absence of contaminants included in the goods using the features that the learning component has acquired,
wherein the inspection irradiator and sensors acquires a final inspection result of the goods on the basis of a plurality of inspection results comprising first inspection results of the goods in a case where it has used the features acquired by the learning machine and second inspection results of the goods in a case where it has not used the features acquired by the learning machine,
the overlapping region data are data relating to regions where the plural articles overlap each other in the inspection images, and
the contaminated region data are data relating to regions where there are contaminants in the inspection images.