Automated inspection data collection for machine learning applications
A method includes collecting measurement data of a sample, determining a confidence value associated with the measurement data, determining if the confidence value is less than a confidence threshold value, and causing the measurement data to be stored in a memory device if the confidence value is less than the confidence threshold value. The method further includes utilizing the measurement data for training of a machine learning system, thereby updating the operation of the machine learning system. The measurement data is measured by an inspection device located at a first location. In one example, the measurement data is a captured image captured by an optical inspector. The confidence value associated with the captured image represents the confidence of the machine learning system regarding what is displayed in the captured image.
1. A method comprising:
(a) manually collecting measurement data of a sample, wherein the manually collected measurement data includes manually generated quality measurement value;
(b) collecting automated measurement data of a sample, wherein the automated measurement data includes an image of the sample and an automatically generated quality measurement value;
(c) determining a difference between the manually generated quality measurement value and the automatically generated quality measurement value;
(d) determining a confidence value based on the difference between the manually generated quality measurement value and the automatically generated quality measurement value; and
(e) causing the automated measurement data to be stored in a memory device if the confidence value is less than the confidence threshold value, wherein the automated measurement data is flagged for machine learning training.
2. The method of claim 1 , further comprising:
(f) utilizing the flagged measurement data for training of a machine learning system, thereby updating the operation of the machine learning system.
3. The method of claim 1 , wherein the automated measurement data is measured by an inspection device located at a first location, and wherein the determining of (d) is performed by a computing device located a second location.
4. The method of claim 1 , wherein the automated measurement data is used to generate the automatically generated quality measurement value.
5. The method of claim 1 , wherein the automated measurement data is measured by an optical inspector.
6. The method of claim 4 , wherein the confidence value determined in (d) represents the confidence of the machine learning system regarding the image of the sample and the automatically generated quality measurement value.
7. The method of claim 1 , wherein the automated measured data is measured utilizing an optical sensor, a moisture sensor, a microtoxin sensor, a thermometer sensor, an acidity (pH) sensor, a microwave sensor, a pressure sensor, a level sensor, an ultrasonic sensor, a flow sensor, a viscosity sensor, a conductance/impedance sensor, an electronic nose (sniffing) sensor, an x-ray sensor, a multi Spectral (visual/non visual) sensor, a weight sensor, a refractometer sensor, a tenderometer sensor, a firmness sensor, a hardness sensors, or a proximity sensor.
8. The method of claim 1 , wherein the automated measured data is collected by an inspection device, wherein the inspection device is an in-flight 3D inspector or an adaptable inspector unit.
9. The method of claim 8 , wherein the inspection device communicates directly with the memory device without passing through a network external to the inspection device.
10. The method of claim 8 , wherein the inspection device communicates with the memory device via an external network.
11. The method of claim 10 , wherein the memory device is part of the computing system located at a second location.
12. The method of claim 11 , wherein the (f) utilizing the measurement data for training of a machine learning system, thereby updating the operation of the machine learning system, is performed by the inspection device.
13. The method of claim 11 , wherein the (f) utilizing the measurement data for training of a machine learning system, thereby updating the operation of the machine learning system, is performed by the computing system located at the second location.
14. The method of claim 13 , wherein the updated operation of the machine learning system is communicated to the inspection device from the computing system located at the second location.
15. The method of claim 3 , wherein the confidence threshold value is determined by the inspection device.
16. The method of claim 11 , wherein the confidence threshold value is determined by the computing system located at the second location.
17. The method of claim 1 , wherein determining the confidence value threshold comprises:
(a) performing steps (a) through (e) for a plurality of samples;
(b) generate a table of confidence values and corresponding manually generated quality measurement values and automatically generated quality measurement values for each of the plurality of samples; and
(c) determine an average confidence value where the difference between the manually generated quality measurement value and automatically generated quality measurement value is greater than fifty percent.
18. The method of claim 17 , wherein the confidence value for each of the plurality of samples are averaged together to calculate an average confidence value threshold.
19. The method of claim 1 , wherein the determining a confidence value associated with the measurement data is performed by the inspection device.
20. The method of claim 1 , wherein the determining a confidence value associated with the measurement data is performed by the computing system located at a second location.