Object sensing device

• US 9,679,196 B2
• Filed: 02/03/2014
• Issued: 06/13/2017
• Est. Priority Date: 02/27/2013
• Status: Active Grant

• Current Assignee: Hitachi Astemo Ltd. (Hitachi, Ltd.)
• Original Assignee: Hitachi Automotive Systems Limited (Hitachi, Ltd.)
• Inventors: Matono, Haruki, Kakegawa, Shinji, Shima, Takeshi, Otsuka, Yuji
• Primary Examiner(s): Azarian, Seyed
• Application Number:
  US14/770,604
• Publication Number:
  US 20160012282A1
• Time in Patent Office: 1,226 Days
• Field of Search: 382100, 382103-107, 382154-155, 382162, 382173, 382181, 382189-194, 382199, 382209, 382219, 382232, 382243, 382254, 382274, 382276, 382285-291, 382305, 382312, 701301, 701 36, 348135, 180272
• CPC Class Codes: B60R 1/24 in front of the vehicle
• CPC Class Codes: B60R 1/31 providing stereoscopic vision
• CPC Class Codes: B60R 2300/30 characterised by the type of image processing
• CPC Class Codes: B60R 2300/60 characterised by monitoring and displaying vehicle exterior scenes from a transformed perspective
• CPC Class Codes: B60R 2300/804 for lane monitoring
• CPC Class Codes: G06T 2207/10021 Stereoscopic video; Stereoscopic image sequence
• CPC Class Codes: G06T 2207/20076 Probabilistic image processing
• CPC Class Codes: G06T 2207/30196 Human being; Person
• CPC Class Codes: G06T 2207/30252 Vehicle exterior; Vicinity of vehicle
• CPC Class Codes: G06T 2207/30256 Lane; Road marking
• CPC Class Codes: G06T 7/593 from stereo images
• CPC Class Codes: G06V 10/96 Management of image or video recognition tasks
• CPC Class Codes: G06V 20/35 Categorising the entire scene, e.g. birthday party or wedding scene
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle
• 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 40/103 Static body considered as a whole, e.g. static pedestrian or occupant recognition
• CPC Class Codes: H04N 13/239 using two 2D image sensors having a relative position equal to or related to the interocular distance H04N13/243 takes precedence
• CPC Class Codes: H04N 2013/0092 Image segmentation from stereoscopic image signals
• CPC Class Codes: H04N 23/73 by influencing the exposure time
• CPC Class Codes: H04N 7/188 Capturing isolated or intermittent images triggered by the occurrence of a predetermined event, e.g. an object reaching a predetermined position signal generation from motion picture films H04N5/253

Abstract

In order to provide an object sensing device whereby, among limited computation resources, performance is improved in sensing an object when sense processing a plurality of objects to be sensed, an object sensing device includes image capture units which capture images external to a host vehicle, and a processing device which sense processes objects to be sensed from the images which are captured by the image capture units, said processing device further including: a scene analysis unit which analyzes a travel scene of the host vehicle; a process priority change unit which changes a sensing process priority of the object to be sensed on the basis of the travel scene which is analyzed by the scene analysis unit; and an object to be sensed sensing unit which carries out a sensing of the object to be sensed on the basis of the sensing process priority which is changed by the process priority change unit.

Claims

1. An object sensing device, comprising:

an image capture unit that captures surroundings of a host vehicle; and

at least one processor that executes a sensing process of an object to be sensed from an image captured by the image capture unit to;

analyze a travel scene of the host vehicle;

change a sensing process priority of the object to be sensed based on the travel scene analyzed by the scene analysis unit;

sense the object to be sensed based on the changed sensing process priority,

calculate an existence probability of the object to be sensed based on map information and the analyzed travel scene using pre-stored learned data in which travel scene types, place attributes, and existence probability values are associated.

2. The object sensing device according to claim 1, wherein the image capture unit comprises a left image capture unit and a right image capture unit, and

the at least one processor executes a sensing process of the object to be sensed from a left image captured by the left image capture unit and a right image captured by the right image capture unit.

3. The object sensing device according to claim 2, wherein the at least one processor:

calculates a distance to the object to be sensed from the left image and the right image; and

calculates an existence probability of the object to be sensed based on the travel scene analyzed by the scene analysis unit,

wherein the at least one processor changes the sensing process priority of the object to be sensed based on the calculated existence probability, and

executes a sensing process of the object to be sensed based on calculated distance information and the changed sensing process priority.

4. The object sensing device according to claim 3, wherein the at least one processor:

extracts a road region based on the left image, the right image, and the distance information;

detects a parked vehicle based on the left image, the right image, and the distance information;

determines attributes of places in regions outside of the road region based on the left image, the right image, and the distance information; and

detects a crosswalk based on the left image, the right image, and the distance information.

5. The object sensing device according to claim 1, wherein the at least one processor calculates an existence probability of the object to be sensed based on the analyzed travel scene, and

changes the sensing process priority of the object to be sensed based on the calculated existence probability.

6. The object sensing device according to claim 5, wherein the at least one processor generates and outputs a speed command for controlling a speed of the host vehicle based on the calculated existence probability.

7. The object sensing device according to claim 5, wherein the at least one processor changes exposure control parameters of the image capture unit based on the calculated existence probability.

8. The object sensing device according to claim 7, wherein if the object to be sensed is a pedestrian, the at least one processor:

extracts regions in which a possibility that a pedestrian exists is high from the calculated existence probability;

changes a pedestrian sensing logic that senses a pedestrian based on the regions in which a possibility that a pedestrian exists is high;

performs gradation correction on an image based on the regions in which a possibility that a pedestrian exists is high; and

extracts a brightness of an image from the regions in which a possibility that a pedestrian exists is high and changes exposure control parameters based on the brightness of the image.

9. The object sensing device according to claim 1, wherein the at least one processor:

acquires the map information, and

wherein the at least one processor changes the sensing process priority of the object to be sensed based on the calculated existence probability.

10. The object sensing device according to claim 1, wherein the image capture unit and the at least one processor are installed in an integrated housing or are installed in separate housings.

11. The object sensing device according to claim 1, wherein if a pedestrian existence probability is higher than a predetermined value, the at least one processor changes the sensing process priority so that a pedestrian sensing process is executed with priority over other objects to be detected.

12. The object sensing device according to claim 11, wherein the at least one processor changes sensing process parameters so that the pedestrian sensing process when the pedestrian existence probability is high is executed in detail.

13. The object sensing device according to claim 12, wherein if the pedestrian existence probability is higher than the predetermined value, a command for executing speed control by suppressing acceleration of the host vehicle is generated and output to a vehicle speed control device.

14. The object sensing device according to claim 13, wherein the pedestrian existence probability is calculated based on learned data, the learned data prepared based on one or more results of investigations conducted in advance.

15. The object sensing device according to claim 14, wherein the learned data is included in a table, wherein a vertical axis of the table includes image scene types and a horizontal axis includes one or more attributes places where the host vehicle is traveling.

16. The object sensing device according to claim 15, wherein the image scene types include one or more of:

(i) guard rail or shrubbery, (ii) gaps between guard rail or shrubbery, (iii) crosswalks, (iv) sidewalks without guard rail, (v) parked vehicles, and (vi) gaps between parked vehicles, and wherein the one or more attributes places includes one or more of;

(i) an urban area, (ii) a residential area, (iii) a commercial facility, (iv) a school, (v) highway, (vi) an elevated road, (vii) a mountainous area, and (viii) a road with few intersections.

17. The object sensing device according to claim 1, wherein the at least one processor changes sensing process parameters so that a pedestrian sensing process when a pedestrian existence probability is high is executed in detail.

18. The object sensing device according to claim 1, wherein if a pedestrian existence probability is higher than a predetermined value, a command for executing speed control by suppressing acceleration of the host vehicle is generated and output to a vehicle speed control device.

19. The object sensing device according to claim 1, wherein a pedestrian existence probability is calculated based on learned data, the learned data prepared based on one or more results of investigations conducted in advance.

20. The object sensing device according to claim 19, wherein the learned data is included in a table, wherein a vertical axis of the table includes image scene types and a horizontal axis includes one or more attributes places where the host vehicle is traveling.

21. The object sensing device according to claim 20, wherein the image scene types include one or more of:

(i) guard rail or shrubbery, (ii) gaps between guard rail or shrubbery, (iii) crosswalks, (iv) sidewalks without guard rail, (v) parked vehicles, and (vi) gaps between parked vehicles, and wherein the one or more attributes places includes one or more of;

(i) an urban area, (ii) a residential area, (iii) a commercial facility, (iv) a school, (v) highway, (vi) an elevated road, (vii) a mountainous area, and (viii) a road with few intersections.