Back to results
US 12,586,313 B2Granted
Damage detection from multi-view visual data
AssigneeFyusion, Inc.
InventorsStefan Johannes Josef Holzer, Abhishek Kar, Matteo Munaro, Pavel Hanchar, Radu Bogdan Rusu, Santi Arano
Life of this patent
Estimating the term…
Abstract
A plurality of images may be analyzed to determine an object model. The object model may have a plurality of components, and each of the images may correspond with one or more of the components. Component condition information may be determined for one or more of the components based on the images. The component condition information may indicate damage incurred by the object portion corresponding with the component.
Claims20
A method comprising:
- determining an object model by applying a neural network to a first plurality of images, each of the first plurality of images being captured from a respective viewpoint, the object model including a plurality of object model components, each of the images corresponding to one or more of the object model components, each of the object model components corresponding to a respective portion of the designated object, wherein the object model is associated with calibration data;
- determining respective component condition information for one or more of the object model components using the plurality of images, the component condition information indicating a characteristic of damage incurred by the respective object portion corresponding with the object model component, wherein a statistical estimate of damage triggers capture of additional images;
- converting the characteristic of damage into a heatmap visual representation associated with estimated damage to the respective object portion, wherein the heatmap shows the probability and the severity of damage to the respective object portion;
- overlaying the heatmap onto a Multi-View Interactive Digital Media Representation (MVIDMR) associated with the designated object;
- and
- storing the component condition information on a storage device.
3 more elements
One or more non-transitory computer readable media having instructions stored thereon for performing a method, the method comprising:
- determining an object model by applying a neural network to a first plurality of images, each of the first plurality of images being captured from a respective viewpoint, the object model including a plurality of object model components, each of the images corresponding to one or more of the object model components, each of the object model components corresponding to a respective portion of the designated object, wherein the object model is associated with calibration data;
- determining respective component condition information for one or more of the object model components using the plurality of images, the component condition information indicating a characteristic of damage incurred by the respective object portion corresponding with the object model component, wherein a statistical estimate of damage triggers capture of additional images;
- converting the characteristic of damage into a heatmap visual representation associated with estimated damage to the respective object portion, wherein the heatmap shows the probability and the severity of damage to the respective object portion;
- overlaying the heatmap onto a Multi-View Interactive Digital Media Representation (MVIDMR) associated with the designated object;
- and
- storing the component condition information on a storage device.
3 more elements
A mobile computing device comprising:
- a camera configured to capture a video of a designated object as it moves through space around the designated object;
- a processor configured to:
- determine an object model by applying a neural network to a first plurality of images, each of the first plurality of images being captured from a respective viewpoint, the object model including a plurality of object model components, each of the images corresponding to one or more of the object model components, each of the object model components corresponding to a respective portion of the designated object, wherein the object model is associated with calibration data;
- determine respective component condition information for one or more of the object model components using the plurality of images, the component condition information indicating a characteristic of damage incurred by the respective object portion corresponding with the object model component, wherein a statistical estimate of damage triggers capture of additional images;
- convert the characteristic of damage into a heatmap visual representation associated with estimated damage to the respective object portion, wherein the heatmap shows the probability and the severity of damage to the respective object portion;
- and
- overlay the heatmap onto a Multi-View Interactive Digital Media Representation (MVIDMR) associated with the designated object;
- and
- a storage device configured to store the component condition information.
6 more elements
Dependencies are inferred from the claim text.