VEHICLE SYSTEMS, DEVICES, AND METHODS FOR RECOGNIZING EXTERNAL WORLDS

• US 20130073194A1
• Filed: 08/02/2012
• Published: 03/21/2013
• Est. Priority Date: 09/15/2011
• Status: Abandoned Application

• Current Assignee: Clarion Co., Ltd. (Faurecia SA)
• Original Assignee: Clarion Co., Ltd. (Faurecia SA)
• Inventors: NAKAMURA, Katsuyuki, YOSHINAGA, Tomoaki, MORINAGA, Mitsutoshi, OGATA, Takehito
• Application Number:
  US13/565,335
• Publication Number:
  US 20130073194A1
• Time in Patent Office: Days
• US Class Current: 701/301
• CPC Class Codes: G06F 18/254 of classification results, e.g. of results related to same input data
• CPC Class Codes: G06V 10/22 by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
• CPC Class Codes: G06V 10/809 of classification results, e.g. where the classifiers operate on the same input data
• 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 20/588 Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road

Abstract

Preferably, an object such as a vehicle is detected regardless of a distance up to the object. A device for recognizing external worlds that analyzes an image acquired by capturing the vicinity of a self-vehicle includes: a processing area setting unit setting a first area of the image indicating a short range and a second area of the image indicating a long range, a first object detecting unit detecting the object by means of a first classifier in the set first area, a second object detecting unit detecting the object by considering even a background pattern by means of a second classifier in the set second area, a rectangular correction unit correcting a detected object rectangular shape and a time to collision (TTC) computing unit computing a prediction time up to a collision based on the detected object rectangular shape.

Claims

1. A method for recognizing external worlds by an external world recognizing device that analyzes a captured image and detects an object, wherein the external world recognizing device sets a first area and a second area for detecting the object in the image, and detects the object by using both an object pattern and a background pattern of the corresponding object pattern at the time of detecting the object in the set second area.

2. The method for recognizing external worlds according to claim 1, wherein:

the first area is an area of a shorter range than the second area, and

the external world recognizing device detects the object by using only the object pattern at the time of detecting the object in the first area.

3. The method for recognizing external worlds according to claim 1, wherein the external world recognizing device corrects an object rectangular shape including the background pattern detected in the second area as an object rectangular shape without the corresponding background pattern.

4. The method for recognizing external worlds according to claim 3, wherein the external world recognizing device computes a prediction time up to a collision with an object corresponding to the corresponding object rectangular shape by using the object rectangular shape detected in the first area or the object rectangular shape after the correction.

5. The method for recognizing external worlds according to claim 4, wherein:

the object is a vehicle, and

the external world recognizing device generates a vehicle width of the vehicle from the object rectangular shape detected in the first area or the object rectangular shape after the correction, and computes the prediction time based on the vehicle width.

6. A device for recognizing external worlds that analyzes a captured image and detects an object, comprising:

a processing area setting unit setting a first area and a second area for detecting the object in the image; and

first and second object detecting units detecting the objects in the set first area and second area, respectively,

wherein the first object detecting unit uses only an object pattern at the time of detecting the object and the second object detecting unit uses both the object pattern and a background pattern of the corresponding object pattern at the time of detecting the object.

7. The device for recognizing external worlds according to claim 6, wherein:

the object is a vehicle, and

the first area is an area of a shorter range than the second area.

8. The device for recognizing external worlds according to claim 7, wherein:

the first object detecting unit and the second object detecting unit include a first classifier and a second classifier, respectively, and

the first classifier is constituted by a plurality of weak classifiers for determining a back-surface pattern of the vehicle and the second classifier is constituted by a plurality of weak classifiers for determining the back-surface pattern of the vehicle and the background pattern.

9. The device for recognizing external worlds according to claim 8, wherein the second object detecting unit includes a rectangular correction unit correcting an object rectangular shape including the background pattern detected in the second area as an object rectangular shape without the corresponding background pattern.

10. The device for recognizing external worlds according to claim 9, further comprising:

a time to collision (TTC) computing unit computing a prediction time up to a collision with an object corresponding to the corresponding object rectangular shape by using the object rectangular shape detected by the first object detecting unit or the object rectangular shape after the correction corrected by the rectangular correction unit.

11. The device for recognizing external worlds according to claim 10, wherein the time to collision (TTC) computing unit generates a vehicle width of the vehicle by using the object rectangular shape detected by the first object detecting unit or the object rectangular shape after the correction corrected by the rectangular correction unit, and computes the prediction time based on the vehicle width.

12. A vehicle system including an external world recognizing device that detects a vehicle by analyzing an image acquired by capturing the vicinity of a self-vehicle, wherein:

the external world recognizing device includes a processing unit and a storage unit,

the storage unit stores a first classifier and a second classifier, and

the processing unit

sets a first area for detecting a vehicle and a second area of a longer range than the first area, in the image,

detects a vehicle rectangular shape of the vehicle by determining a vehicle pattern by means of the first classifier, in the first area,

detects the vehicle rectangular shape of the vehicle by determining the vehicle pattern and a background pattern of the corresponding vehicle pattern by means of the second classifier, in the second area,

corrects the vehicle rectangular shape detected in the second area, and

computes a time to collision (TTC) up to a collision with the self-vehicle based on the vehicle rectangular shape detected by using the first classifier or the vehicle rectangular shape detected and corrected by using the second classifier.

13. The vehicle system according to claim 12, wherein the processing unit sets the first area and the second area based on detection of a lane in the image.

14. The vehicle system according to claim 12, wherein the processing unit sets the first area and the second area based on a yaw rate.

15. The vehicle system according to claim 12, wherein the processing unit computes a collision risk in which the self-vehicle collides with the vehicle in accordance with the collision prediction time, and performs a control for avoiding the collision of the self-vehicle in accordance with the computed collision risk.