Image processing device and computer-readable storage medium
An image processing device includes: a model storage unit storing a model for estimating, from image data, a first probability map indicating feature points of a workpiece in the image data, and a second probability map indicating areas between the feature points; an estimation unit estimating a first probability map and a second probability map on the basis of the image data using the model stored in the model storage unit; a feature point estimation unit estimating feature points of the same workpiece in the image data on the basis of the first probability map and the second probability map; a position and orientation estimation unit estimating the position and orientation of the workpiece in the image data on the basis of the result of the estimation by the feature point estimation unit; and an output unit outputting the result of the estimation by the position and orientation estimation unit.
1 . An image processing device for generating a machine learning model used to estimate a position and attitude of a workpiece from within image data, on the basis of the image data obtained by imaging the workpiece, the image processing device comprising:
a data acquisition unit configured to acquire image data obtained by imaging a workpiece;
a preprocessing unit configured to create a first probability map indicating feature points of the workpiece in the image data and a second probability map indicating regions between the feature points on the basis of the image data acquired by the data acquisition unit;
a learning unit configured to generate a model of machine learning for estimating the first probability map and the second probability map from the image data on the basis of the image data, the first probability map, and the second probability map; and
a model storage unit configured to store the model created by the learning unit.
2 . An image processing device for estimating a position and attitude of a workpiece from within image data obtained by imaging the workpiece, the image processing device comprising:
a model storage unit configured to store a model for estimating a first probability map indicating feature points of a workpiece in image data and a second probability map indicating regions between the feature points from the image data obtained by imaging the workpiece;
a data acquisition unit configured to acquire image data obtained by imaging at least one workpiece;
an estimation unit configured to estimate a first probability map and a second probability map, using the model stored in the model storage unit, on the basis of the image data acquired by the data acquisition unit;
a feature point estimation unit configured to estimate feature points of the same workpiece in the image data on the basis of the first probability map and the second probability map;
a position and attitude estimation unit configured to estimate a position and attitude of the workpiece in the image data on the basis of a result estimated by the feature point estimation unit; and
an output unit configured to output a result estimated by the position and attitude estimation unit.
3 . A non-transitory storage medium storing a program causing a computer to function as an image processing device for generating a machine learning model used to estimate a position and attitude of a workpiece in image data obtained by imaging the workpiece on the basis of the image data, the program causing the computer to function as:
a data acquisition unit configured to acquire image data obtained by imaging a workpiece;
a preprocessing unit configured to create a first probability map indicating feature points of the workpiece in the image data and a second probability map indicating regions between the feature points on the basis of the image data acquired by the data acquisition unit;
a learning unit configured to generate a machine learning model for estimating the first probability map and the second probability map from the image data on the basis of the image data, the first probability map, and the second probability map; and
a model storage unit configured to store the model created by the learning unit.
4 . A non-transitory storage medium storing a program causing a computer to function as an image processing device for estimating a position and attitude of a workpiece in image data obtained by imaging the workpiece, the program causing the computer to function as:
a model storage unit configured to store a model for estimating a first probability map indicating feature points of a workpiece in image data and a second probability map indicating regions between the feature points from the image data obtained by imaging the workpiece;
a data acquisition unit configured to acquire image data obtained by imaging at least one workpiece;
an estimation unit configured to estimate a first probability map and a second probability map, using the model stored in the model storage unit, on the basis of the image data acquired by the data acquisition unit;
a feature point estimation unit configured to estimate feature points of the same workpiece in the image data on the basis of the first probability map and the second probability map;
a position and attitude estimation unit configured to estimate a position and attitude of the workpiece in the image data on the basis of a result estimated by the feature point estimation unit; and
an output unit configured to output a result estimated by the position and attitude estimation unit.