| CPC G06Q 40/08 (2013.01) | 9 Claims |
|
1. A computer-implemented method comprising:
causing a remote user device to display instructions that indicate how a plurality of acceptable images of a vehicle, and for determining by a machine learning algorithm configured to determine a severity of damage to a vehicle and a cost to repair damage, a severity of damage to a vehicle and a cost to repair damage to the vehicle, should be captured, wherein receiving the plurality of acceptable images comprises iteratively, and until a sufficient quantity of acceptable images is received:
receiving, via the remote device, one or more images, wherein at least one of the one or more images corresponds to a three-dimensional image;
generating, via image processing logic and object recognition logic, image metadata associated with the one or more images, wherein the image metadata specifies at least one of blur data, depth data, angle data, or object data, wherein the object data specifies at least one identified object and a corresponding location of the at least one identified object within each image;
determining whether each of the one or more images is acceptable for use with the machine learning algorithm, wherein determining whether each image of the plurality of images is acceptable for use in the machine learning algorithm comprises:
comparing the image data of a particular image of the plurality of images to a predefined set of criteria to assess a blurriness of the particular image; and
when the assessed blurriness of the particular image is beyond a threshold blurriness, determine the particular image to be unacceptable for determining the severity of damage on the vehicle;
adding acceptable images of the one or more images to the plurality of acceptable images; and
after determining that one or more of the one or more images are unacceptable images, causing the remote user device to display instructions that indicate how to capture images related to damage to the vehicle, wherein the determination causes the remote user device to display a user interface screen including feedback regarding changes to be made for capturing one or more replacement images to replace the one or more of the one or more unacceptable images;
after receiving the sufficient quantity of acceptable images, determining, based on one or more acceptable images of the plurality of acceptable images, damage information comprising a location of damage on the vehicle and an indication of severity of damage on the vehicle, wherein the indication of severity of damage of the vehicle is made using stereoscopic image processing techniques and machine learning to classify dents and determine depth of dents on the vehicle; and
determining a repair estimate for the damage to the vehicle.
|