Systems and methods for object detection
Systems and methods for detecting objects from a moving platform are provided. The method includes receiving, from a camera coupled to the moving platform, first and second image frames, wherein the second image frame is subsequent in time to the first image frame. Within the first image frame, a first target is defined characterizing the location of an object detected in the first image frame. A first score associated with the first target and a second score associated with the second target are determined, and a first predicted target is determined based on a tracking process applied to the first target. A third score associated with the first predicted target is determined. The method determines a merged target corresponding to the second target when the second score is greater than the third score and a proximity measure between the second target and the first predicted target is above a threshold.
1. A method for detecting objects from a moving platform, the method comprising:
receiving, from a camera coupled to the moving platform, a first image frame;
defining, within the first image frame, a first target characterizing a location of an object detected in the first image frame;
determining a first detection score associated with the first target;
receiving, from the camera, a second image frame subsequent in time to the first image frame;
defining, within the second image frame, a second target characterizing the location of the object in the second image frame;
determining a second detection score associated with the second target;
defining, within the second image frame, a first predicted target based on a tracking process applied to the first target;
determining a third detection score associated with the first predicted target; and
defining a merged target corresponding to the second target when the second detection score is greater than the third detection score and a proximity measure between the second target and the first predicted target is above a predetermined threshold.
2. The method of claim 1 , further comprising defining the merged target corresponding to the first predicted target when the third detection score is greater than the second detection score and the overlap area of the second target and the first predicted target is greater than a predetermined threshold.
3. The method of claim 1 , wherein the moving platform is a vehicle.
4. The method of claim 1 , further comprising increasing the merged target detection score based on local motion detected within at least one of the first image frame and the second image frame.
5. The method of claim 4 , further comprising decreasing the merged target detection score when a velocity of the moving platform is above a threshold value.
6. The method of claim 1 , further comprising ignoring the second target when it has not been detected for a predetermined number of frames.
7. The method of claim 1 , further comprising modifying the merged target detection score based on a distance from the camera to an object corresponding to the merged target.
8. The method of claim 1 , further comprising modifying the merged target detection score based on a track history associated with the merged target.
9. An object detection system for a moving platform, the system comprising:
a visual tracking module; and
a detection module communicatively coupled to the visual tracking module, the detection module configured to:
receive, from a camera coupled to the moving platform, a first image frame and a second image frame subsequent in time to the second image frame;
define, within the first image frame, a first target characterizing a location of an object detected in the first image frame and to define, within the second image frame, a second target characterizing the location of the object in the second image frame;
determine a first detection score associated with the first target;
determine a second detection score associated with the second target;
receive, from the visual tracking module, a first predicted target based on a tracking process applied to the first target;
determine a third detection score associated with the first predicted target; and
define a merged target corresponding to the second target when the second detection score is greater than the third detection score and the overlap area of the second target and the first predicted target is greater than a predetermined threshold.
10. The system of claim 9 , wherein the detection module is further configured to define the merged target corresponding to the first predicted target when the third detection score is greater than the second detection score and the overlap area of the second target and the first predicted target is greater than a predetermined threshold.
11. The system of claim 9 , wherein the moving platform is a vehicle.
12. The system of claim 9 , further comprising a target motion module configured to increase the merged target detection score based on local motion detected within at least one of the first image frame and the second image frame.
13. The system of claim 12 , wherein the target motion module is configured to decrease the merged target detection score when a velocity of the moving platform is above a threshold value.
14. The system of claim 9 , further including ignoring the second target when it has not been detected for a predetermined number of frames.
15. The system of claim 9 , wherein the detection module is configured to modify the merged target detection score based on a distance from the camera to an object corresponding to the merged target.
16. The system of claim 9 , wherein the detection module is configured to modify the merged target detection score based on a track history associated with the merged target.
17. A vehicle, comprising:
a camera configured to produce a plurality of image frames associated with an environment in the vicinity of the vehicle;
an object detection system communicatively coupled to the camera, the system comprising:
a visual tracking module; and
a detection module communicatively coupled to the visual tracking module, the detection module configured to:
receive, from a camera coupled to the moving platform, a first image frame and a second image frame subsequent in time to the second image frame;
define, within the first image frame, a first target characterizing a location of an object detected in the first image frame and to define, within the second image frame, a second target characterizing the location of the object in the second image frame;
determine a first detection score associated with the first target;
determine a second detection score associated with the second target;
receive, from the visual tracking module, a first predicted target based on a tracking process applied to the first target;
determine a third detection score associated with the first predicted target; and
define a merged target corresponding to the second target when the second detection score is greater than the third detection score and the overlap area of the second target and the first predicted target is greater than a predetermined threshold.
18. The vehicle of claim 17 , wherein the detection module is further configured to define the merged target corresponding to the first predicted target when the third detection score is greater than the second detection score and the overlap area of the second target and the first predicted target is greater than a predetermined threshold.
19. The vehicle of claim 17 , wherein the object detection system includes a target motion module configured to increase the merged target detection score based on local motion detected within at least one of the first image frame and the second image frame.
20. The vehicle of claim 17 , wherein the detection module is configured to modify the merged target detection score based on a distance from the camera to an object corresponding to the merged target.