Systems and methods for determining physical parameters of feet

Abstract

Methods, systems, and non-transitory computer readable media for computing physical dimensions of feet based on user-captured images are described. In at least one embodiment, an exemplary method comprises: receiving, by a server from a user device, an image of the user's foot or feet; segmenting the image to identify the user's foot or feet; computing the one or more physical parameters of the user's foot or feet.

What is claimed is: 
     
         1 . A method of computing one or more physical parameters of a user's foot or feet, the method comprising:
 receiving, by a server from a user device, an image of the user's foot or feet;   segmenting the image to identify the user's foot or feet and a fiducial marker present in the image; and   computing the one or more physical parameters of the user's foot or feet based on at least one known dimension of the identified fiducial marker.   
     
     
         2 . The method of  claim 1 , wherein the image is the sole image received from the user device that is used to compute the one or more physical parameters of the user's foot or feet. 
     
     
         3 . The method of  claim 1 , wherein the image is received in response to a single user action to capture the image by the user device. 
     
     
         4 . The method of  claim 1 , wherein one of the one or more physical parameters include a recommended shoe size and/or a foot width. 
     
     
         5 . The method of  claim 1 , further comprising:
 transmitting the one or more physical parameters to the user device to be presented to the user.   
     
     
         6 . The method of  claim 1 , wherein the image was captured by a camera of the user device when the direction of the camera was oriented substantially along a normal direction of a surface on which the user's foot or feet were positioned. 
     
     
         7 . The method of  claim 1 , wherein segmenting the image to identify the user's foot or feet comprises:
 for each foot present in the image, applying a deep learning model to detect a centroid of the foot.   
     
     
         8 . The method of  claim 1 , wherein segmenting the image to identify the user's foot or feet comprises:
 for each foot present in the image, applying one or more deep learning models to:
 segment a foot region within the image; 
 segment a toe region of the foot; 
 segment a heel region of the foot; and 
 generate a combined segmented region from the foot region, the toe region, and the heel region. 
   
     
     
         9 . The method of  claim 1 , wherein the segmenting is performed using a region-based convolutional neural network. 
     
     
         10 . A method of computing one or more physical parameters of a user's foot or feet, the method comprising:
 receiving, by a server from a user device, an image of the user's foot or feet and depth map data associated with the image;   segmenting the image to identify the user's foot or feet; and   computing the one or more physical parameters of the user's foot or feet based on the depth map data.   
     
     
         11 . The method of  claim 10 , wherein the depth map data corresponds to data obtained via a LiDAR sensor of the user device. 
     
     
         12 . The method of  claim 10 , wherein computing the one or more physical parameters of the user's foot or feet based on the depth map data comprises:
 detecting, within three-dimensional (3D) point cloud corresponding to the depth map data, a plane representative of a surface on which the user's feet are positioned in the image;   generating a 2D projection of the foot or feet by projecting the 3D point cloud onto the plane based on the user's foot or feet identified by the segmentation; and   computing a scaling parameter from the 2D projection that is used to compute the one or more physical parameters.   
     
     
         13 . The method of  claim 10 , wherein the image is the sole image received from the user device that is used to compute the one or more physical parameters of the user's foot or feet. 
     
     
         14 . The method of  claim 10 , wherein the image is received in response to a single user action to capture the image by the user device. 
     
     
         15 . The method of  claim 10 , wherein one of the one or more physical parameters include a recommended shoe size and/or a foot width. 
     
     
         16 . The method of  claim 10 , wherein the image was captured by a camera of the user device when the direction of the camera was oriented substantially along a normal direction of a surface on which the user's foot or feet were positioned. 
     
     
         17 . The method of  claim 10 , wherein segmenting the image to identify the user's foot or feet comprises:
 for each foot present in the image, applying a deep learning model to detect a centroid of the foot.   
     
     
         18 . The method of  claim 10 , wherein segmenting the image to identify the user's foot or feet comprises:
 for each foot present in the image, applying one or more deep learning models to:
 segment a foot region within the image; 
 segment a toe region of the foot; 
 segment a heel region of the foot; and 
 generate a combined segmented region from the foot region, the toe region, and the heel region. 
   
     
     
         19 . The method of  claim 10 , wherein the segmenting is performed using a region-based convolutional neural network. 
     
     
         20 . A method of computing one or more physical parameters of a user's foot or feet, the method comprising:
 generating, for display by a user device, a video feed captured by a camera of the user device;   generating an overlay on the video feed, the overlay comprising:
 a visual indicator of where the user is to position their foot or feet on a surface; and 
 optionally, a visual indicator of where the user is to position a fiducial marker; 
   receiving a user input as a result of a single user action, wherein responsive to the user input, the user device captures one or more images of the user's feet and, optionally, the fiducial marker;   transmitting the one or more images to a server, wherein the server is to compute one or more physical dimensions of the user's foot or feet from the one or more images; and   presenting for display the one or more physical dimensions of the user's foot or feet received from the server.