VEHICLE WITH IMPROVED TRAFFIC-OBJECT POSITION DETECTION

• US 20130335569A1
• Filed: 03/13/2013
• Published: 12/19/2013
• Est. Priority Date: 03/14/2012
• Status: Active Grant

• Current Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Original Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Inventors: EINECKE, Nils, REBHAN, Sven, NISHIGAKI, Morimichi
• Granted Patent: US 9,313,462 B2
• Time in Patent Office: Days
• US Class Current: 348/148
• CPC Class Codes: G01S 13/42 Simultaneous measurement of distance and other co-ordinates indirect measurement G01S13/46
• CPC Class Codes: G01S 13/58 Velocity or trajectory determination systems; Sense-of-movement determination systems
• CPC Class Codes: G01S 13/726 Multiple target tracking
• CPC Class Codes: G01S 13/867 Combination of radar systems with cameras
• CPC Class Codes: G01S 13/931 of land vehicles
• CPC Class Codes: G01S 2013/932 using own vehicle data, e.g. ground speed, steering wheel direction
• CPC Class Codes: G01S 2013/9323 Alternative operation using light waves
• CPC Class Codes: G01S 2013/9324 Alternative operation using ultrasonic waves
• CPC Class Codes: G01S 2013/93271 in the front of the vehicles
• CPC Class Codes: G01S 7/2955 Means for determining the position of the radar coordinate system for evaluating the position data of the target in another coordinate system G01S7/24 takes precedence; sighting devices adapted for indirect laying of fire F41G3/16; inertial navigation G01C21/16
• CPC Class Codes: G06V 20/58 Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
• CPC Class Codes: G06V 20/584 of vehicle lights or traffic lights
• CPC Class Codes: G08G 1/166 for active traffic, e.g. moving vehicles, pedestrians, bikes
• CPC Class Codes: H04N 7/18 Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast

Abstract

A method is proposed for detecting the position of a target object in the environment of a host vehicle, comprising the steps of

• • receiving a first position of the target object as detected by an object position sensor like a radar sensor;
  • receiving an image containing the target object as detected by an image sensor like a camera;
  • projecting the first position on the image; and
  • refining the first position by computing a second position of the target object on the basis of a symmetry search within the image.

Claims

1. A method for detecting the position of a target object in the environment of a host vehicle, comprising the following steps:

receiving a first position information on the target object using an object position sensor like e.g. a radar sensor,

receiving an image containing the target object as detected by an image sensor like e.g. a camera,

projecting the first position on the image, and

refining the first position by computing a second refined position of the target object on the basis of a symmetry search within the image.

2. The method according to claim 1,

wherein the first position and the second position are lateral positions of the target object preferably reflecting the angle or the direction to the target vehicle with respect to the host vehicle.

3. The method according to claim 1,

wherein the symmetry search consists in identifying a symmetry axis of the detected image of the target object.

4. The method of claim 1,

wherein the symmetry search is performed within a preferably horizontal symmetry search area, the symmetry search area being located in the image in accordance with the first position and being preferably centered on the first position.

5. The method according to claim 4,

wherein a symmetry window, which size preferably corresponds to the size of the target object, is positioned within the symmetry search area, and

a correlation value is computed between the content of the symmetry window and the mirrored content of the symmetry window.

6. The method of claim 5,

wherein a correlation value is computed for different positions of the symmetry window within symmetry search area, and

at least one peak is detected among the computed correlation values.

7. The method according to claim 4,

wherein the symmetry search is performed by horizontally folding the symmetry window with its mirrored image using a locally normalizing cost function as a similarity measure for calculating a symmetry value for each position of the symmetry window.

8. The method according to claim 7,

wherein the locally normalizing cost function is the summed normalized cross-correlation (SNCC), local mean and variance normalized (MVN) images with sum of absolute difference (SAD), rank transformed (RT) images with sum of absolute difference (SAD) or census transformed images with Hamming distance.

9. The method according to claim 4,

wherein the symmetry search is performed in frequency domain, e.g. according to phase only correlation (POC), to compute horizontal shifts between the symmetry window and its mirrored image.

10. The method according to claim 4,

wherein the symmetry search area is defined by the distance between host vehicle and target object as detected by the object position sensor, and by an average physical width for the target object.

11. The method according to claim 4,

wherein the symmetry search area is down sampled to a pre-fixed resolution so that the maximum computational costs per image is bound only by the number of objects and not their distance.

12. The method of claim 1,

wherein the first position is refined only if the target object is identified as a non-static object, or alternatively the first position is refined depending on the relevance of the target object to the host vehicle, e.g. the first position is refined only if the target object is identified as belonging to the own and/or neighboring lanes.

13. The method of claim 1,

wherein it is evaluated if the target object is partly or fully occluded in the image, preferably by projecting onto the image the position of other objects detected by the object position sensor and evaluating potential occlusion by means of the distance of the hypotheses to the host vehicle.

14. The method of claim 1,

wherein the symmetry search is governed by at least one consistency check for preventing an erroneous refining of the first position.

15. The method according to claim 14,

wherein a consistency check evaluates if the symmetry search leads to multiple valid solutions, e.g. to multiple symmetry peaks, for the second position.

16. The method according to claim 14,

wherein a symmetry peak obtained via the symmetry search is not taken into account for refining the first position if the symmetry peak is below a given threshold, or if the region around the symmetry peak and above the given threshold is wider than a give fraction of the expected width of the target object.

17. The method according to claim 14,

wherein detected symmetry peaks have to further pass a consistency check that evaluates the past position of theses peaks.

18. The method of claim 1,

wherein the position of a detected symmetry peak is stabilized by computing a locally weighted mean around the symmetry peak.

19. The method of claim 1,

wherein the second position is incorporated into a Kalman filter for tracking the detected position over time for a spatio-temporal stabilization.

20. The method of claim 1,

wherein the first position or the second position is incorporated into a Kalman filter, the noise parameters of the Kalman filter being dynamically adapted to a low noise configuration in case the second position is processed by the Kalman filter, or to a higher noise configuration in case the first position is processed by the Kalman filter e.g. because the position refining failed.

21. A device for detecting the position of a target object in the environment of a host vehicle, the device comprising:

means for receiving a first position of the target object as detected by an object position sensor like e.g. a radar sensor,

means for receiving an image containing the target object as detected by an image sensor like e.g. a camera,

means for projecting the first position on the image, and

means for refining the first position by computing a second position of the target object on the basis of a symmetry search within the image.

22. A system for detecting the position of a target object in the environment of a host vehicle, the system comprising:

an object position sensor, for example a radar sensor, for detecting a first position of the target object,

an image sensor, for example a camera, for detecting an image containing the target object,

a device for projecting the first position on the image, and for refining the first position by computing a second position of the target object on the basis of a symmetry search within the image.

23. A land, air or sea vehicle, preferably a car or motorcycle, comprising the system of claim 22.