On-board radar apparatus and region detection method

• US 10,101,448 B2
• Filed: 11/24/2015
• Issued: 10/16/2018
• Est. Priority Date: 12/12/2014
• Status: Active Grant

• Current Assignee: Panasonic Automotive Systems Company of America (Panasonic Holdings Corporation)
• Original Assignee: Panasonic Intellectual Property Management Co., Ltd. (Panasonic Holdings Corporation)
• Inventors: Hamada, Asako, Nishimura, Hirofumi, Kobayashi, Kiyotaka, Shikatani, Maiko
• Primary Examiner(s): Bythrow, Peter M
• Assistant Examiner(s): Seraydaryan, Helena H
• Application Number:
  US14/951,481
• Publication Number:
  US 20160170020A1
• Time in Patent Office: 1,057 Days
• Field of Search: 342 70
• CPC Class Codes: G01S 13/66 Radar-tracking systems; Analogous systems
• CPC Class Codes: G01S 13/723 by using numerical data
• 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 2013/93271 in the front of the vehicles
• CPC Class Codes: G01S 7/414 Discriminating targets with respect to background clutter

Abstract

An on-board radar apparatus includes a transmitter/receiver that transmits a radar signal to a detection range for every frame and receives one or more reflected signals which are the radar signal reflected by one or more objects; a detector that detects, for every frame in each direction within the detection range, a position of a reflection point closest to the on-board radar apparatus as a boundary candidate position, which serves as a boundary with a region where no object exists within the detection range, among one or more reflection points detected on the basis of the one or more reflected signals; a calculator that calculates movement amount concerning an amount of movement of the on-board radar apparatus; an estimator that generates, for every frame in each direction within the detection range, an estimated boundary position by converting the boundary candidate position detected in a past frame into a boundary position in a current frame on the basis of the movement amount; and a smoother that performs, for every frame in each direction within the detection range, smoothing processing by using the boundary candidate position in the current frame and the estimated boundary position to calculate a boundary position with the region where no object exists within the detection range and outputs the calculated the boundary position to a driving support apparatus.

Claims

1. An on-board radar apparatus comprising:

a transmitter/receiver, which, in operation, transmits a radar signal to a detection range in every frame and receives one or more reflected signals which are the radar signal reflected by one or more objects;

a detector, which, in operation, detects, for every frame in each direction within the detection range, a position of a reflection point closest to the on-board radar apparatus as a boundary candidate position, for a region where no object exists within the detection range, among one or more reflection points detected in the respective frame for the one or more reflected signals;

a calculator, which, in operation, calculates movement amount concerning an amount of movement of the on-board radar apparatus;

an estimator, which, in operation, generates, for every frame in each direction within the detection range, an estimated boundary position from the boundary candidate position detected in a past frame and a conversion into a current frame on the basis of the movement amount; and

a smoother, which, in operation, performs, for every frame in each direction within the detection range, smoothing processing by using the boundary candidate position in the current frame and the estimated boundary position for the current frame to determine a presence or absence of a boundary position with the region where no object exists within the detection range, and outputs the boundary position to a driving support apparatus in a case where the determined result is the presence of the boundary position.

2. The on-board radar apparatus according to claim 1, wherein the smoother calculates, for every frame in each direction within the detection range, a likelihood value of the boundary candidate position or the estimated boundary position by using a preset likelihood weight, determines, for every frame in each direction within the detection range, that the boundary position is present in a case where the likelihood value is equal to or larger than a predetermined threshold value, and outputs, in a case where the determined result is the presence of the boundary position in each direction within the detection range, for every frame in each direction within the detection range, the boundary position calculated by using a weighted average of distances between the on-board radar apparatus and the boundary candidate position in the current frame or the estimated boundary position using preset output value weights.

3. The on-board radar apparatus according to claim 1, wherein

the one or more objects includes two or more objects, and the two or more objects include a still object and a moving object.

4. The on-board radar apparatus according to claim 1, wherein

the one or more objects are one or more still objects.

5. The on-board radar apparatus according to claim 1, further comprising:

a camera, which, in operation, captures an image of the detection range;

a region setter, which, in operation, sets an image region used for object recognition in the captured image on the basis of the boundary position; and

an object recognizer, which, in operation, performs object recognition on the set image region.

6. A region detection method comprising:

transmitting a radar signal to a detection range in every frame;

receiving one or more reflected signals which are the radar signal reflected by one or more objects;

detecting, for every frame in each direction within the detection range, a position of a reflection point closest to the on-board radar apparatus as a boundary candidate position, for a region where no object exists within the detection range, among one or more reflection points detected in the respective frame for the one or more reflected signals;

calculating movement amount data concerning an amount of movement of the on-board radar apparatus;

generating, for every frame in each direction within the detection range, an estimated boundary position from the boundary candidate position detected in a past frame and conversion into a current frame on the basis of the movement amount data;

performing, for every frame in each direction within the detection range, smoothing processing by using the boundary candidate position in the current frame and the estimated boundary position for the current frame to determine a presence or absence of a boundary position with the region where no object exists within the detection range; and

outputting, for every frame in each direction within the detection range, the boundary position to a driving support apparatus in a case where the determined result is the presence of the boundary position.

7. The region detection method according to claim 6, wherein

the performing of the smoothing processing includes:

calculating, for every frame in each direction within the detection range, a likelihood value of the boundary candidate position or the estimated boundary position by using a preset likelihood weight; and

determining, for every frame in each direction within the detection range, that the boundary position is present in a case where the likelihood value is equal to or larger than a predetermined threshold value, and

in the outputting, in a case where the determined result is the presence of the boundary position in each direction within the detection range, for every frame in each direction within the detection range, the boundary position calculated by using a weighted average of distances between the on-board radar apparatus and the boundary candidate position in the current frame or the estimated boundary position, using preset output value weights.

8. The region detection method according to claim 6, wherein

the one or more objects includes two or more objects and the two or more objects include a still object and a moving object.

9. The region detection method according to claim 6, wherein

the one or more objects are one or more still objects.

10. The region detection method according to claim 6, further comprising:

capturing an image of the detection range by using a camera;

setting an image region used for object recognition in the captured image on the basis of the boundary position; and

performing object recognition on the set image region.