Semantic label based filtering of objects in an image generated from high definition map data
The autonomous vehicle generates an overlapped image by overlaying HD map data over sensor data and rendering the overlaid images. The visualization process is repeated as the vehicle drives along the route. The visualization may be displayed on a screen within the vehicle or at a remote device. The system performs reverse rendering of a scene based on map data from a selected point. For each line of sight originating at the selected point, the system identifies the farthest object in the map data. Accordingly, the system eliminates objects obstructing the view of the farthest objects in the HD map as viewed from the selected point. The system further allows filtering of objects using filtering criteria based on semantic labels. The system generates a view from the selected point such that 3D objects matching the filtering criteria are eliminated from the view.
1. One or more non-transitory computer readable storage media storing instructions encoded thereon that, in response to being executed by one or more processors cause a system to perform operations comprising:
accessing map data from a map datastore, the map datastore storing metadata describing one or more real-world objects with semantic labels, wherein the map data corresponds to a scene surrounding a current location of a vehicle;
for at least one angle in a plurality of angles originating at a point of origin associated with the vehicle, selecting a point on a first object from the map data, the point being closest to the point of origin along a line through the map data at the angle, the first object having a semantic label that fails to match a semantic label specified by a filtering criterion;
rendering an image of a view from the point of origin based on the selected points;
transmitting the generated image for display on a client device; and
overlaying the points of the first object over a second object, the second object being positioned between the first object and the point of origin and having a semantic label matching a semantic label specified in the filtering criterion,
wherein the second object is shown in the generated image in a lighter shade than the first object.
2. The one or non-transitory computer readable storage media of claim 1 , wherein the filtering criterion identifies a type of real-world objects.
3. The one or non-transitory computer readable storage media of claim 1 , wherein the filtering criterion identifies a particular object in the map data.
4. The one or non-transitory computer readable storage media of claim 1 , wherein the generated image shows objects corresponding to unselected points as outlines.
5. The one or non-transitory computer readable storage media of claim 1 , wherein the operations further comprise:
transmitting the generated image to a remote display device for displaying a video, the video showing the scene as observed from the point of view of the autonomous vehicle as the autonomous vehicle moves.
6. The one or non-transitory computer readable storage media of claim 1 , wherein the operations further comprise:
dynamically updating the generated image as the autonomous vehicle moves.
7. A computer-implemented method comprising:
accessing map data from a map datastore, the map datastore storing metadata describing one or more real-world objects with semantic labels, wherein the map data corresponds to a scene surrounding a current location of a vehicle;
for at least one angle in a plurality of angles originating at a point of origin associated with the vehicle, selecting a point on a first object from the map data, the point being closest to the point of origin along a line through the map data at the angle, the first object having a semantic label that fails to match a semantic label specified by a filtering criterion;
rendering an image of a view from the point of origin based on the selected points;
transmitting the generated image for display on a client device; and
overlaying the points of the first object over a second object, the second object being positioned between the first object and the point of origin and having a semantic label matching a semantic label specified in the filtering criterion,
wherein the second object is shown in the generated image in a lighter shade than the first object.
8. The computer-implemented method of claim 7 , wherein the filtering criterion identifies a type of real-world objects.
9. The computer-implemented method of claim 7 , wherein the filtering criterion identifies a particular object in the map data.
10. The computer-implemented method of claim 7 , wherein the generated image shows objects corresponding to unselected points as outlines.
11. The computer-implemented method of claim 7 , further comprising:
transmitting the generated image to a remote display device for displaying a video, the video showing the scene as observed from the point of view of the autonomous vehicle as the autonomous vehicle moves.
12. The computer-implemented method of claim 7 , further comprising:
dynamically updating the generated image as the autonomous vehicle moves.
13. A computer system comprising:
one or more processors; and
one or more non-transitory computer readable storage media storing instructions encoded thereon that, in response to being executed by the one or more processors, cause the system to perform operations comprising:
accessing map data from a map datastore, the map datastore storing metadata describing one or more real-world objects with semantic labels, wherein the map data corresponds to a scene surrounding a current location of a vehicle;
for at least one angle in a plurality of angles originating at a point of origin associated with the vehicle, selecting a point on a first object from the map data, the point being closest to the point of origin along a line through the map data at the angle, the first object having a semantic label that fails to match a semantic label specified by a filtering criterion;
rendering an image of a view from the point of origin based on the selected points;
transmitting the generated image for display on a client device; and
overlaying the points of the first object over a second object, the second object being positioned between the first object and the point of origin and having a semantic label matching a semantic label specified in the filtering criterion,
wherein the second object is shown in the generated image in a lighter shade than the first object.
14. The computer system of claim 13 , wherein the filtering criterion identifies a type of real-world objects.
15. The computer system of claim 13 , wherein the filtering criterion identifies a particular object in the map data.
16. The computer system of claim 13 , wherein the generated image shows objects corresponding to unselected points as outlines.
17. The computer system of claim 13 , wherein the operations further comprise:
transmitting the generated image to a remote display device for displaying a video, the video showing the scene as observed from the point of view of the autonomous vehicle as the autonomous vehicle moves.
18. The computer system of claim 13 , wherein the operations further comprise:
dynamically updating the generated image as the autonomous vehicle moves.