Teacher data generation method, trained learning model, and system
According to one aspect of the present invention, a method of generating teacher data for image recognition includes acquiring image data by capturing an image of a workpiece, and segmenting the image data into a plurality of first areas, marking whether predetermined information is included in each of the plurality of first areas, and generating a plurality of pieces of teacher data.
1. A system comprising:
an image capturing apparatus;
one or more processors; and
one or more memories storing instructions and a trained learning model configured to receive image data obtained by capturing an image by the image capturing apparatus,
wherein the trained learning model was created
by inputting difference information indicating a difference between first image data obtained by capturing an image of a first workpiece and second image data different from the first image data, wherein the first image data and the second image data were obtained by capturing one or more images of a first workpiece model, wherein the first workpiece is an instance of the first workpiece model, and
by inputting difference information indicating a difference between third image data obtained by capturing an image of a second workpiece and fourth image data different from the third image data, wherein the third image data and the fourth image data were obtained by capturing one or more images of a second workpiece model that is different from the first workpiece model, wherein the second workpiece is an instance of the second workpiece model,
wherein executing the instructions causes the one or more processors and the one or more memories to input difference information indicating a difference between fifth image data obtained by capturing an image of a third workpiece by the image capturing apparatus and sixth image data different from the fifth image data to the trained learning model and make a determination on the third workpiece,
wherein the difference information indicating the difference between the fifth image data and the sixth image data to be input to the learning model indicates information subjected to predetermined processing, and
wherein in a case where an absolute value of a difference value in an area including the difference indicated by the difference information between the fifth image data and the sixth image data exceeds a predetermined threshold, the predetermined processing is processing of correcting the area to an area including no difference.
2. The system according to claim 1 , wherein the second image data, the fourth image data, and the sixth image data are each image data obtained by capturing an image of a respective non-defective workpiece.
3. The system according to claim 1 ,
wherein the first workpiece includes a repetitive pattern,
wherein the second image data indicates a shift image obtained by shifting the first image data by an amount corresponding to one or more repetitive patterns in image data obtained by capturing an image of the first workpiece,
wherein the fourth image data indicates a shift image obtained by shifting the third image data by an amount corresponding to one or more repetitive patterns in image data obtained by capturing an image of the second workpiece, and
wherein the sixth image data indicates a shift image obtained by shifting the fifth image data by an amount corresponding to one or more repetitive patterns in image data obtained by capturing an image of the third workpiece.
4. The system according to claim 1 , wherein the determination indicates whether a defect is present.
5. The system according to claim 1 ,
wherein the third workpiece is a colored part, and
wherein the determination indicates whether the third workpiece is colored with a color different from a predetermined color.
6. The system according to claim 1 , further comprising:
a sensor; and
a trigger generation circuit configured to transmit an image capturing trigger signal to the image capturing apparatus,
wherein upon detection of the third workpiece within a predetermined range, the sensor outputs a signal to the trigger generation circuit, and
wherein the image capturing apparatus captures an image of the third workpiece based on the image capturing trigger signal output from the trigger generation circuit based on the signal.
7. The system according to claim 6 , further comprising a robot,
wherein executing the instructions causes the one or more processors and the one or more memories to cause a workpiece determined to be defective based on the trained learning model to move.
8. A system comprising:
an image capturing apparatus;
one or more processors; and
one or more memories storing instructions and a trained learning model configured to receive image data obtained by capturing an image by the image capturing apparatus,
wherein the trained learning model was created by inputting difference information indicating a difference between first image data obtained by capturing an image of a first workpiece by the image capturing apparatus and second image data different from the first image data, and
wherein executing the instructions causes the one or more processors and the one or more memories to
generate difference information indicating at least one difference between third image data obtained by capturing an image of a second workpiece by the image capturing apparatus and fourth image data different from the third image data, wherein in a case where an absolute value of a difference value in an area including a difference between the third image data and the fourth image data exceeds a predetermined threshold, the difference information indicates that the area includes no difference, and
input the difference information to the trained learning model and make a determination on the second workpiece.