Systems and methods using image recognition processes for improved operation of a laundry appliance
A method may include obtaining one or more images of the washing machine appliance from a camera assembly of a remote device and detecting a fiducial reference on a front portion of the washing machine appliance within the one or more images. The method may also include comparing the detected fiducial reference to a two-dimensional reference shape in an obtained image of the images and determining a set camera angle for the camera assembly is met based on the comparison in the obtained image. The method may further include analyzing the obtained image using a machine learning image recognition process and estimating a load attribute of a load of clothes within the washing machine appliance based on the analysis. The method may still further include directing a wash cycle within the washing machine appliance based on the estimated load attribute.
1. A method of operating a washing machine appliance, the washing machine appliance comprising a cabinet, a wash tub, and a wash basket, the wash tub being mounted within the cabinet, and the wash basket being rotatably mounted within a wash tub and defining a wash chamber configured for receiving a load of clothes, the method comprising:
obtaining one or more images of the washing machine appliance from a camera assembly of a remote device spaced apart from the cabinet;
detecting a fiducial reference on a front portion of the washing machine appliance within the one or more images;
comparing the detected fiducial reference to a two-dimensional reference shape in an obtained image of the one or more images;
determining a set camera angle for the camera assembly is met based on the comparison in the obtained image;
analyzing the obtained image using a machine learning image recognition process to estimate a load attribute of a load of clothes within the washing machine appliance based on the analysis,
the load attribute comprising at least one of a fabric type, a load color, or a load size; and
directing a wash cycle such that the wash cycle is performed within the washing machine appliance based on the estimated load attribute.
2. The method of claim 1 , wherein the washing machine appliance defines an opening to the wash chamber to permit access thereto, and wherein the fiducial reference is the opening.
3. The method of claim 1 , wherein the two-dimensional reference shape is a circle.
4. The method of claim 1 , further comprising:
determining a door of the washing machine appliance is open prior to determining the set camera angle is met.
5. The method of claim 1 , further comprising:
determining a door of the washing machine appliance is closed within a predetermined time period following obtaining one or more images of the washing machine appliance,
wherein directing the wash cycle is in response to determining the door is closed within the predetermined time period.
6. The method of claim 1 , further comprising:
generating a feedback signal prompting a feedback action at the remote device in response to determining the set camera angle is met.
7. The method of claim 1 , wherein obtaining one or more images comprises receiving a video signal from the camera assembly, and wherein the method further comprises
presenting a real-time feed of the camera assembly at the remote device according to the received video signal; and
overlaying the two-dimensional reference shape over the real-time feed.
8. The method of claim 1 , wherein obtaining one or more images comprises receiving a video signal from the camera assembly, and wherein the method further comprises
presenting a real-time feed of the camera assembly at the remote device according to the received video signal; and
displaying movement guidance with the real-time feed to align the two-dimensional reference shape with the fiducial reference.
9. The method of claim 1 , wherein the machine learning image recognition process comprises at least one of a convolution neural network (“CNN”), a region-based convolution neural network (“R-CNN”), a deep belief network (“DBN”), a deep neural network (“DNN”), or a vision transformer (“ViT”) image recognition process.
10. A method of operating a washing machine appliance, the washing machine appliance comprising a cabinet, a wash tub, and a wash basket, the wash tub being mounted within the cabinet, and the wash basket being rotatably mounted within a wash tub and defining a wash chamber configured for receiving a load of clothes, the method comprising:
obtaining one or more images of the washing machine appliance from a camera assembly of a remote device spaced apart from the cabinet, obtaining one or more images comprising receiving a video signal from the camera assembly;
presenting a real-time feed of the camera assembly at the remote device according to the received video signal;
detecting a fiducial reference on a front portion of the washing machine appliance within the one or more images;
comparing the detected fiducial reference to a two-dimensional reference shape in an obtained image of the one or more images;
overlaying the two-dimensional reference shape over the real-time feed;
determining a set camera angle for the camera assembly is met based on the comparison in the obtained image;
analyzing the obtained image using a machine learning image recognition process to estimate a load attribute of a load of clothes within the washing machine appliance based on the analysis,
the load attribute comprising at least one of a fabric type, a load color, or a load size; and
directing a wash cycle such that the wash cycle is performed within the washing machine appliance based on the estimated load attribute.
11. The method of claim 10 , wherein the washing machine appliance defines an opening to the wash chamber to permit access thereto, and wherein the fiducial reference is the opening.
12. The method of claim 10 , wherein the two-dimensional reference shape is a circle.
13. The method of claim 10 , further comprising:
determining a door of the washing machine appliance is open prior to determining the set camera angle is met.
14. The method of claim 10 , further comprising:
determining a door of the washing machine appliance is closed within a predetermined time period following obtaining one or more images of the washing machine appliance,
wherein directing the wash cycle is in response to determining the door is closed within the predetermined time period.
15. The method of claim 10 , further comprising:
generating a feedback signal prompting a feedback action at the remote device in response to determining the set camera angle is met.
16. The method of claim 10 , wherein the method further comprises
displaying movement guidance with the real-time feed to align the two-dimensional reference shape with the fiducial reference.
17. The method of claim 10 , wherein the machine learning image recognition process comprises at least one of a convolution neural network (“CNN”), a region-based convolution neural network (“R-CNN”), a deep belief network (“DBN”), a deep neural network (“DNN”), or a vision transformer (“ViT”) image recognition process.