Vehicle undercarriage imaging
Images of an undercarriage of a vehicle may be captured via one or more cameras. A point cloud may be determined based on the images. The point cloud may includes points positioned in a virtual three-dimensional space. A stitched image may be determined based on the point cloud by projecting the point cloud onto a virtual camera view. The stitched image may be stored on a storage device.
1. A method comprising:
capturing a plurality of images of an undercarriage of a vehicle, the images captured via one or more cameras, each camera being arranged in a respective fixed position, the images being captured as the vehicle moves over the one or more cameras;
determining a designated point cloud based on the plurality of images via a processor, the designated point cloud including a plurality of points positioned in a virtual three-dimensional space;
determining a designated stitched image based on the designated point cloud via the processor, wherein the stitched image is determined by projecting the designated point cloud onto a virtual camera view, wherein determining the stitched image comprises identifying a plurality of correspondence points, each of the correspondence points indicating a visual feature included in two or more of the plurality of images, wherein determining the stitched image further comprises identifying a plurality of point correspondence transformations, each of the point correspondence transformations identifying a respective spatial relationship between a first location of a designated one of the correspondence points in a first one of the images and a second location of the designated correspondence point in a second one of the images; and
storing the designated stitched image on a storage device.
2. The method recited in claim 1 , wherein the virtual camera view is a two-dimensional plane in the virtual three-dimensional space.
3. The method recited in claim 1 , wherein the virtual camera view is a point in the virtual three-dimensional space.
4. The method recited in claim 1 , wherein determining the designated point cloud comprises determining a plurality of point clouds.
5. The method recited in claim 4 , wherein the one or more cameras includes a plurality of cameras, and wherein each of the point clouds corresponds with a respective one of the cameras.
6. The method recited in claim 4 , wherein each of the point clouds corresponds with a respective one of the images.
7. The method recited in claim 1 , wherein the designated point cloud is determined based at least in part on depth information captured from one or more depth sensors.
8. The method recited in claim 1 , wherein determining the stitched image further comprises determining an image correspondence providing a mathematical transformation between the first and second images in a two or three dimensional coordinate space based on the plurality of correspondence transformations.
9. The method recited in claim 8 , wherein the image correspondence is a numerical average of one or more of the point correspondence transformations, the numerical average determined based on a cost function.
10. The method recited in claim 1 , wherein the designated stitched image is one of a plurality of stitched images determined based on the designated point cloud.
11. The method recited in claim 1 , the method further comprising generating a multi-view interactive digital media representation that includes the plurality of stitched images, the multi-view interactive digital media representation being navigable in one or more dimensions.
12. The method recited in claim 11 , wherein the multi-view interactive digital media representation being navigable in three dimensions.
13. A system comprising:
one or more cameras configured to capture a plurality of images of an undercarriage of a vehicle, each camera being arranged in a respective fixed position, the images being captured as the vehicle moves over the one or more cameras;
a processor configured to determine a designated point cloud based on the plurality of images and determine a designated stitched image based on the designated point cloud, the designated point cloud including a plurality of points positioned in a virtual three-dimensional space, wherein the stitched image is determined by projecting the designated point cloud onto a virtual camera view, wherein determining the stitched image comprises identifying a plurality of correspondence points, each of the correspondence points indicating a visual feature included in two or more of the plurality of images, wherein determining the stitched image further comprises identifying a plurality of point correspondence transformations, each of the point correspondence transformations identifying a respective spatial relationship between a first location of a designated one of the correspondence points in a first one of the images and a second location of the designated correspondence point in a second one of the images; and
a storage device configured to store the designated stitched image.
14. The system recited in claim 13 , wherein determining the stitched image further comprises determining an image correspondence providing a mathematical transformation between the first and second images in a two or three dimensional coordinate space based on the plurality of correspondence transformations, and wherein the image correspondence is a numerical average of one or more of the point correspondence transformations, the numerical average determined based on a cost function.
15. The system recited in claim 13 , wherein the processor is further operable to generate a multi-view interactive digital media representation that includes the plurality of stitched images, the multi-view interactive digital media representation being navigable in one or more dimensions.
16. The system recited in claim 13 , wherein determining the designated point cloud comprises determining a plurality of point clouds, and wherein the one or more cameras includes a plurality of cameras, and wherein each of the point clouds corresponds with a respective one of the cameras.
17. The method recited in claim 1 , wherein the virtual camera view is a two-dimensional plane of a point in the virtual three-dimensional space.
18. One or more non-transitory computer readable media having instructions stored thereon for performing a method, the method comprising:
capturing a plurality of images of an undercarriage of a vehicle, the images captured via one or more cameras, each camera being arranged in a respective fixed position, the images being captured as the vehicle moves over the one or more cameras;
determining a designated point cloud based on the plurality of images via a processor, the designated point cloud including a plurality of points positioned in a virtual three-dimensional space;
determining a designated stitched image based on the designated point cloud via the processor, wherein the stitched image is determined by projecting the designated point cloud onto a virtual camera view, wherein determining the stitched image comprises identifying a plurality of correspondence points, each of the correspondence points indicating a visual feature included in two or more of the plurality of images, wherein determining the stitched image further comprises identifying a plurality of point correspondence transformations, each of the point correspondence transformations identifying a respective spatial relationship between a first location of a designated one of the correspondence points in a first one of the images and a second location of the designated correspondence point in a second one of the images; and
storing the designated stitched image on a storage device.