Image-based spoilage sensing refrigerator
Disclosed are various embodiments for determining whether food in a refrigerator has spoiled. An application executing on a computing device receives an image from a refrigerator of a compartment of the refrigerator. The application then analyzes the image to identify a food item in the image. Subsequently, the application analyzes the image to identify a potentially spoiled area of the food item. The application then compares the potentially spoiled area of the food item to a model of a spoiled food item. Afterwards, the application determines that the identified food item has spoiled based at least in part on a comparison of the potentially spoiled area of the identified food item to the model of the spoiled food item.
1. A system, comprising:
a computing device comprising a processor and a memory; and
an application stored in the memory that, when executed by the processor, causes the computing device to at least:
receive an image of a compartment of a refrigerator, wherein the image is captured using an infrared camera located within the refrigerator;
perform a first round of edge detection on the image to identify a shape of a produce item in the image;
determine a produce item type by comparing the shape of the produce item with a model shape of a respective produce item;
perform a second round of edge detection on the image to identify a potentially bruised area of the produce item;
compare the potentially bruised area of the produce item to a bruised model of the produce item type; and
determine that the produce item has bruised based at least in part on the comparison of the potentially bruised area of the produce item to the bruised model of the produce item type.
2. The system of claim 1 , wherein the application is further configured to send to a client device a notification comprising an indication that the produce item has been bruised.
3. The system of claim 2 , wherein the application is further configured to:
highlight the produce item in the image, thereby generating a modified image of the compartment of the refrigerator; and
include the modified image of the compartment of the refrigerator in the notification.
4. A system, comprising:
a computing device comprising a processor and a memory; and
an application stored in the memory that, when executed by the processor, causes the computing device to at least:
receive an image from a refrigerator of a compartment of the refrigerator;
perform edge detection on the image to identify a shape of a food item in the image;
determine a food item type by comparing the shape of the food item with a model shape of a respective food item;
analyze the image to identify a potentially spoiled area of the food item;
compare the potentially spoiled area of the food item to a spoiled model of the food item type; and
determine that the food item has spoiled based at least in part on the comparison of the potentially spoiled area of the food item to the spoiled model of the food item type.
5. The system of claim 4 , wherein the application is further configured to send to a client device a notification comprising an indication that the food item has spoiled.
6. The system of claim 4 , wherein the image comprises an infrared image generated with an infrared camera located in the refrigerator.
7. The system of claim 4 , wherein the image comprises a visible spectrum image generated with a visible spectrum camera located in the refrigerator.
8. The system of claim 4 , wherein the spoiled model of the food item type comprises a model of mold growing on an exterior of the food item type.
9. The system of claim 4 , wherein the spoiled model of the spoiled food item comprises a model of a subsurface bruise of the food item type.
10. The system of claim 4 , wherein the spoiled model of the food item type comprises a three-dimensional model of the food item type.
11. The system of claim 4 , wherein the spoiled model of the food item type comprises a model of an oxidized surface of the food item type.
12. A computer-implemented method, comprising:
receiving an image from a refrigerator of a compartment of the refrigerator;
performing edge detection on the image to identify a shape of a food item in the image;
determining a food item type by comparing the shape of the food item with a model shape of a respective food item:
analyzing the image to identify a potentially spoiled area of the food item;
comparing the potentially spoiled area of the food item to a spoiled model of the food item type; and
determining that the food item has spoiled based at least in part on the comparison of the potentially spoiled area of the food item to the spoiled model of the food item type.
13. The computer-implemented method of claim 12 , further comprising sending to a client device a notification comprising an indication that the food item has spoiled.
14. The computer-implemented method of claim 12 , wherein the image comprises an infrared image generated with an infrared camera located in the refrigerator.
15. The computer-implemented method of claim 12 , wherein the image comprises a visible spectrum image generated with a visible spectrum camera located in the refrigerator.
16. The computer-implemented method of claim 12 , wherein the spoiled model of the food item type comprises a model of mold growing on an exterior of the food item type.
17. The computer-implemented method of claim 12 , wherein the spoiled model of the food item type comprises a model of a subsurface bruise of the food item type.
18. The computer-implemented method of claim 12 , wherein the spoiled model of the food item comprises a three-dimensional model of the food item type.
19. The computer-implemented method of claim 12 , wherein the spoiled model of the food item comprises a model of an oxidized surface of the food item type.
20. The computer-implemented method of claim 12 , wherein analyzing the image to identify the potentially spoiled area of the food item further comprises determining a percentage of a visible surface of the food item in the image is greater than a threshold.