Multispectral nondestructive characterization of edible objects
View Patent ↗A system for characterizing an object that includes an inedible exterior portion and a potentially edible interior portion comprises a computer system and a plurality of sensors configured to acquire data concurrently about the object. The plurality of sensors include a radar system configured to transmit a radar beam to irradiate the object, and to detect a return signal of the radar beam. The computer system is configured to form a radar image of the interior portion of the object based on the return signal, and to characterizing the object based on the acquired data, by applying a machine learning algorithm to the radar image.
1. A system comprising:
a plurality of sensors, including:
a radar system, configured to acquire data concurrently about an object that includes an inedible exterior portion and a potentially edible interior portion, the radar system being configured to transmit a radar beam from the radar system to irradiate the object, and to detect a return signal of the radar beam; and
an optical sensor configured to acquire optical data of the object; and
a computer system coupled to the plurality of sensors and configured to use the optical data to localize the object relative to an aperture of the radar system and to calculate a trajectory of the object in relation to the aperture of the radar system and form a radar image of the interior portion of the object based on the return signal, and to characterize the object based on the acquired data, wherein characterizing the object includes applying a machine learning algorithm to the radar image to characterize the object.
2. A system as recited in claim 1 , wherein characterizing the object comprises determining whether a defect is present in the object.
3. A system as recited in claim 1 , wherein characterizing the object comprises using the machine learning algorithm to characterize the interior portion of the object.
4. A system as recited in claim 1 , wherein the computer system is further configured to collate data about the object acquired from the plurality of sensors.
5. A system as recited in claim 4 , wherein collating the data comprises using data from a second sensor, of the plurality of sensors, to locate the object in the radar image.
6. A system as recited in claim 1 , wherein the radar system is a millimeter wave or terahertz radar system.
7. A system as recited in claim 1 , wherein the optical sensor comprises a short-wave infrared (SWIR) camara.
8. A system as recited in claim 1 , wherein the optical sensor comprises a stereoscopic human-visible light camara.
9. A system as recited in claim 1 , wherein the plurality of sensors further includes an optical sensor, and wherein the computer system is further configured to use an output of the optical sensor to determine a size of the object.
10. A system as recited in claim 1 , wherein:
the plurality of sensors include a microphone; and
the computer system is further configured to use the microphone to acquire sound data of a sound of the object impacting a surface, and to use the sound data to determine a quality of the object.
11. A system as recited in claim 1 , wherein characterizing the object comprises identifying a presence of mold in the interior portion of the object.
12. A system as recited in claim 1 , wherein characterizing the object comprises identifying a type of mold present in the interior portion of the object.
13. A system as recited in claim 1 , wherein characterizing the object comprises identifying a presence of a worm or insect in the interior portion of the object.
14. A system as recited in claim 1 , wherein characterizing the object comprises identifying a presence of rot in the interior portion of the object.
15. A system comprising:
a millimeter wave or terahertz radar system configured to acquire data about an object that includes an inedible exterior portion and a potentially edible interior portion, the radar system being configured to transmit a radar beam from the radar system to irradiate the object, and to detect a return signal of the radar beam;
an optical sensor to acquire optical data of the object concurrently with the radar system transmitting the radar beam or detecting the return signal; and
a computer system coupled to the plurality of sensors and configured to form a radar image of the interior portion of the object based on the return signal, and to characterize the object based on the acquired data, wherein characterizing the object includes
using the optical data to determine a location of the object in the radar image, and
applying a machine learning algorithm to the radar image, based on the determined location of the object, to detect a defect in the interior portion of the object.
16. A system as recited in claim 15 , wherein the optical sensor comprises a short-wave infrared (SWIR) camara.
17. A system as recited in claim 15 , wherein the optical sensor comprises a stereoscopic human-visible light camara.
18. A system as recited in claim 15 , wherein the computer system is further configured to use an output of the optical sensor to determine a size of the object.
19. A system as recited in claim 15 , wherein the plurality of sensors further include a microphone, and the computer system is further configured to use the microphone to acquire sound data of a sound of the object impacting a surface, and to use the sound data to determine a quality of the object.
20. A system as recited in claim 15 , wherein characterizing the object comprises identifying a presence of mold in the interior portion of the object.
21. A system as recited in claim 15 , wherein characterizing the object comprises identifying a type of mold present in the interior portion of the object.
22. A system as recited in claim 15 , wherein characterizing the object comprises identifying a presence of a worm or insect in the interior portion of the object.
23. A system as recited in claim 15 , wherein characterizing the object comprises identifying a presence of rot in the interior portion of the object.
24. A machine-implemented method comprising:
using a plurality of sensors, including a radar system, to acquire data concurrently about an object that includes an inedible exterior portion and a potentially edible interior portion, wherein using the plurality of sensors includes:
using an optical sensor to acquire optical data of the object,
using the optical data to localize the object relative to an aperture of the radar system and to calculate a trajectory of the object in relation to the aperture of the radar system,
transmitting a radar beam from the radar system to irradiate the object,
detecting a return signal of the radar beam, by the radar system, and
forming a radar image of the interior portion of the object based on the return signal; and
characterizing, by a computer, the object based on the acquired data, wherein the characterizing includes applying a machine learning algorithm to the radar image to characterize the object.
25. A method as recited in claim 24 , wherein characterizing the object comprises determining whether a defect is present in the object.
26. A method as recited in claim 24 , wherein characterizing the object comprises using the machine learning algorithm to characterizing the interior portion of the object.
27. A method as recited in claim 24 , further comprising:
collating data about the object acquired from the plurality of sensors.
28. A method as recited in claim 27 , wherein collating the data comprises using data from a second sensor, of the plurality of sensors, to locate the object in the radar image.
29. A method as recited in claim 24 , wherein the radar system is a millimeter wave or terahertz radar system.
30. A method as recited in claim 24 , wherein the optical sensor comprises a short-wave infrared (SWIR) camara.
31. A method as recited in claim 24 , wherein the optical sensor comprises a stereoscopic human-visible light camara.
32. A method as recited in claim 24 , wherein using the plurality of sensors further includes:
using the optical image data to determine a size of the object.
33. A method as recited in claim 24 , wherein:
the plurality of sensors include a microphone;
using the plurality of sensors comprises using the microphone to acquire sound data of a sound of the object impacting a surface; and
characterizing the object includes using the sound data to determine a quality of the object.
34. A method as recited in claim 24 , wherein characterizing the object comprises identifying a presence of mold in the interior portion of the object.
35. A method as recited in claim 24 , wherein characterizing the object comprises identifying a type of mold present in the interior portion of the object.
36. A method as recited in claim 24 , wherein characterizing the object comprises identifying a presence of a worm or insect in the interior portion of the object.
37. A method as recited in claim 24 , wherein characterizing the object comprises identifying a presence of rot in the interior portion of the object.