Image identification system
Embodiments may relate to a graphical user interface (GUI). The GUI may include a first portion that displays an image related to images of a location. The GUI may also include a second portion that displays an image related to detection and ranging information of the location. The two images may be linked such that an interaction with an object in one portion of the GUI causes changes in the other portion of the GUI. Other embodiments may be described or claimed.
1. A method for providing intuitive user interactions that reduce object identification time and improve object identification accuracy, comprising:
generating a first portion of a graphical user interface (GUI) comprising visible-light camera images of a location around a vehicle, wherein the visible-light camera images are laid side by side to form an image strip, and a subset of the visible-light camera images are visible to a user through the first portion of the GUI based on a vision field of the image strip;
generating a second portion the GUI comprising a map displaying data from a light detection and ranging (LIDAR) system and a directional halo representing the vision field of the image strip;
receiving a user interaction in the first portion of the GUI, wherein the user interaction includes a selection of an object in the first portion of the GUI and an alteration of the vision field of the first portion of the GUI;
altering the directional halo of the second portion of the GUI to include the object within directional halo in the second portion of the GUI and to represent the altered vision field; and
causing the object in the second portion of the GUI to be accentuated.
2. The method of claim 1 , further comprising:
receiving a confirmation from the user that the object is properly identified, classified, or labeled.
3. The method of claim 2 , further comprising:
storing objects confirmed by the user using the GUI in a database as annotated data usable to train machine learning models.
4. The method of claim 1 , further comprising:
generating a third portion of the GUI comprising a list of object labels.
5. The method of claim 4 , wherein each object label comprises one or more of: a unique identifier, an object class, and metadata about a corresponding object.
6. The method of claim 4 , further comprising:
causing an object label corresponding to the object in the third portion of the GUI to be accentuated.
7. The method of claim 1 , wherein:
the alteration of the vision field comprises a rotation of the vision field to center the object in the vision field; and
altering the directional halo comprises rotating the directional halo in the second portion of the GUI to represent the altered vision field.
8. The method of claim 1 , wherein:
the alteration of the vision field comprises a zoom action of the vision field to center the object in the vision field; and
altering the directional halo comprises changing a breadth of the directional halo in the second portion of the GUI to represent the altered vision field.
9. The method of claim 1 , wherein the map further displays data from a radio detection and ranging system.
10. One or more non-transitory computer-readable media comprising instructions for providing intuitive user interactions that reduce object identification time and improve object identification accuracy, that, upon execution of the instructions by one or more processors of a computing device, are to cause the computing device to:
generate a first portion of a graphical user interface (GUI) comprising visible-light camera images of a location around a vehicle, wherein the visible-light camera images are laid side by side to form an image strip, and a subset of the visible-light camera images are visible to a user through the first portion of the GUI based on a vision field of the image strip;
generate a second portion the GUI comprising a map displaying data from a light detection and ranging (LIDAR) system and a directional halo representing the vision field of the image strip;
receive a user interaction in the first portion of the GUI, wherein the user interaction includes a selection of an object in the first portion of the GUI and an alteration of the vision field of the first portion of the GUI;
alter the directional halo of the second portion of the GUI to include the object within directional halo in the second portion of the GUI and to represent the altered vision field; and
cause the object in the second portion of the GUI to be accentuated.
11. The one or more non-transitory computer-readable media of claim 10 , wherein the instructions are to cause the computing device to further:
receive a confirmation from the user that the object is properly identified, classified, or labeled; and
store objects confirmed by the user using the GUI in a database as annotated data usable to train machine learning models.
12. The one or more non-transitory computer-readable media of claim 10 , wherein the instructions are to cause the computing device to further:
generate a third portion of the GUI comprising a list of object labels; and
cause an object label corresponding to the object in the third portion of the GUI to be accentuated.
13. The one or more non-transitory computer-readable media of claim 10 , wherein:
the alteration of the vision field comprises a rotation of the vision field to center the object in the vision field; and
altering the directional halo comprises rotating the directional halo in the second portion of the GUI to represent the altered vision field.
14. The one or more non-transitory computer-readable media of claim 10 , wherein:
the alteration of the vision field comprises a zoom action of the vision field to center the object in the vision field; and
altering the directional halo comprises changing a breadth of the directional halo in the second portion of the GUI to represent the altered vision field.
15. The one or more non-transitory computer-readable media of claim 10 , wherein the map further displays data from a radio detection and ranging system.
16. An object identification system for providing intuitive user interactions that reduce object identification time and improve object identification accuracy, comprising:
one or more processors;
one or more memory devices storing instructions to be executed by the one or more processors; and
one or more display devices in communication with the one or more processors to display a graphical user interface (GUI);
wherein the instructions are to cause the one or more processors to:
generate a first portion of a graphical user interface (GUI) comprising visible-light camera images of a location around a vehicle, wherein the visible-light camera images are laid side by side to form an image strip, and a subset of the visible-light camera images are visible to a user through the first portion of the GUI based on a vision field of the image strip;
generate a second portion the GUI comprising a map displaying data from a light detection and ranging (LIDAR) system and a directional halo representing the vision field of the image strip;
receive a user interaction in the first portion of the GUI, wherein the user interaction includes a selection of an object in the first portion of the GUI and an alteration of the vision field of the first portion of the GUI;
alter the directional halo of the second portion of the GUI to include the object within directional halo in the second portion of the GUI and to represent the altered vision field; and
cause the object in the second portion of the GUI to be accentuated.
17. The object identification system of claim 16 , wherein the instructions are to cause the one or more processors to further:
receive a confirmation from the user that the object is properly identified, classified, or labeled; and
store objects confirmed by the user using the GUI in a database as annotated data usable to train machine learning models.
18. The object identification system of claim 16 , wherein the instructions are to cause the one or more processors to further:
generate a third portion of the GUI comprising a list of object labels; and
cause an object label corresponding to the object in the third portion of the GUI to be accentuated.
19. The object identification system of claim 16 , wherein:
the alteration of the vision field comprises a rotation of the vision field to center the object in the vision field; and
altering the directional halo comprises rotating the directional halo in the second portion of the GUI to represent the altered vision field.
20. The object identification system of claim 16 , wherein:
the alteration of the vision field comprises a zoom action of the vision field to center the object in the vision field; and
altering the directional halo comprises changing a breadth of the directional halo in the second portion of the GUI to represent the altered vision field.