Autonomous vehicle vision system
In accordance with one aspect of the invention, there is provided a method of autonomously operating a vehicle. The method provides at least two cameras in operable communication with the vehicle for providing substantially similar views relative to the vehicle. The at least two cameras receive information relating to the views. The method also provides a laser in operable communication with the vehicle for selectively shining a single discrete mark on at least a portion of the views provided by the at least two cameras. Further, the method determines whether the information received by the at least two cameras is ambiguous regarding the views. The method activates the laser on at least a portion of the views based on whether the information received by the at least two cameras is ambiguous.
1. A method of autonomously operating an autonomous vehicle, comprising:
providing a vision system having at least two cameras in operable communication with the autonomous vehicle for providing substantially similar views relative to the vehicle, the at least two cameras receiving information relating to the views and the views at least being substantially in the direction of travel of the vehicle, the two cameras alone at least capable of determining distances between the autonomous vehicle and any objects in the autonomous vehicle's path;
providing a stand-alone laser in operable communication with the vehicle only for selectively shining a single discrete mark on at least a portion of the views provided to the at least two cameras, wherein a shape of the single discrete mark is determined from a template stored in a memory of the autonomous vehicle;
determining whether the information received by the at least two cameras is at least ambiguous regarding the views and therefore capable of more than one interpretation with regard to the direction of travel of the vehicle;
in response to determining that the information received by the at least two cameras is ambiguous, activating the laser to project the single discrete mark into the views provided by the at least two cameras, wherein the single discrete mark is projected in the shape determined from the template stored in the memory of the autonomous vehicle;
detecting the projected single discrete mark within a corresponding portion of each of the views provided by the at least two cameras, the detecting based on analyzing objects contained in each of the views against expected shape information of the single discrete mark contained within the template;
in response to detecting the projected single discrete mark within each of the views provided by the at least two cameras, deactivating the laser; and
calculating a distance to an object upon which the projected single discrete mark impinges, the calculating based on image data contained within the corresponding portion in which the projected single discrete mark was detected in each of the views provided by the at least two cameras, in order to thereby resolve the ambiguity.
2. The method of claim 1 , wherein providing the at least two cameras comprises providing the at least two cameras for providing overlapping views relative to the vehicle.
3. The method of claim 1 , wherein providing the at least two cameras comprises providing the at least two cameras for providing views relative to the vehicle that overlap by at least a threshold percentage.
4. The method of claim 1 , wherein providing the at least two cameras comprises providing the at least two cameras for providing views relative to the vehicle whose fields of view overlap by at least a threshold angle.
5. The method of claim 1 , wherein providing the laser comprises providing the laser for shining a dot, a square, a diamond, or an irregular shape.
6. The method of claim 1 , wherein providing the laser comprises providing the laser for shining the discrete mark with a curvature.
7. The method of claim 1 , wherein providing the laser comprises providing the laser for shining the discrete mark with a straight edge.
8. The method of claim 1 , wherein providing the laser comprises providing the laser for shining the discrete mark that has a solid interior.
9. The method of claim 1 , wherein determining whether the information is ambiguous comprises:
determining whether a result of a formula when inputs based on the information in the substantially similar views are used in the formula exceeds a predetermined threshold.
10. The method of claim 1 , wherein determining whether the information is ambiguous comprises determining whether a distance to an object in the views can be ascertained.
11. The method of claim 1 , wherein determining whether the information is ambiguous comprises determining whether a distance to an object in the views can be logically determined.
12. The method of claim 1 , wherein activating the laser based on whether the information is ambiguous comprises activating the laser only when the information is ambiguous.
13. The method of claim 1 , wherein activating the laser based on whether the information is ambiguous comprises not activating the laser when the information is not ambiguous.
14. The method of claim 1 wherein activating the laser comprises activating the laser on at least a portion of an object to illuminate the portion to the at least two cameras.
15. The method of claim 14 further comprising using the at least two cameras to determine the distance to the illuminated object.