System and method for evaluating the perception system of an autonomous vehicle

• US 9,679,191 B1
• Filed: 07/07/2015
• Issued: 06/13/2017
• Est. Priority Date: 10/05/2010
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Waymo LLC (Alphabet Inc.)
• Inventors: Urmson, Christopher Paul, Smith, Russell Leigh, Zhu, Jiajun, Fairfield, Nathaniel, Haehnel, Dirk
• Primary Examiner(s): Yang, Weiwen
• Application Number:
  US14/792,995
• Time in Patent Office: 707 Days
• Field of Search: None
• CPC Class Codes: B60R 1/23 with a predetermined field of view
• CPC Class Codes: B60R 2300/30 characterised by the type of image processing
• CPC Class Codes: B60T 17/18 Safety devices; Monitoring
• CPC Class Codes: B60T 17/221 Procedure or apparatus for checking or keeping in a correct functioning condition of brake systems hydraulic pressure systems in general F15B19/00, F15B21/04; testing structures or apparatus G01M
• CPC Class Codes: B60T 2201/022 Collision avoidance systems
• CPC Class Codes: B60T 2210/32 Vehicle surroundings
• CPC Class Codes: B60T 2270/406 Test-mode; Self-diagnosis
• CPC Class Codes: B60T 7/22 initiated by contact of vehicle, e.g. bumper, with an external object, e.g. another vehicle , or by means of contactless obstacle detectors mounted on the vehicle
• CPC Class Codes: B60T 8/00 Arrangements for adjusting wheel-braking force to meet varying vehicular or ground-surface conditions, e.g. limiting or varying distribution of braking force by changing number of effective brake cylinders in power brake systems B60T17/10
• CPC Class Codes: B60T 8/17 Using electrical or electronic regulation means to control braking detecting or indicating faulty operation B60T8/885
• CPC Class Codes: B60T 8/885 using electrical circuitry
• CPC Class Codes: B60W 2050/0292 Fail-safe or redundant systems, e.g. limp-home or backup systems
• CPC Class Codes: B60W 2420/403 Image sensing, e.g. optical camera
• CPC Class Codes: B60W 2420/408 Radar; Laser, e.g. lidar
• CPC Class Codes: B60W 2552/05 Type of road, e.g. motorways, local streets, paved or unpaved roads
• CPC Class Codes: B60W 2555/60 Traffic rules, e.g. speed limits or right of way
• CPC Class Codes: B60W 2556/10 Historical data
• CPC Class Codes: B60W 2556/50 of positioning data, e.g. GPS [Global Positioning System] data
• CPC Class Codes: B60W 30/08 Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
• CPC Class Codes: B60W 30/186 excessive wear or burn out of friction elements, e.g. clutches
• CPC Class Codes: B60W 40/06 : Road conditions
• CPC Class Codes: B60W 50/14 : Means for informing the driver, warning the driver or prompting a driver intervention
• CPC Class Codes: B62D 6/00 : Arrangements for automatically controlling steering depending on driving conditions sensed and responded to, e.g. control circuits means for initiating a change in direction B62D1/00; steering valves B62D5/06; combined with means for inclining the vehicle body or wheels on bends B62D9/00
• CPC Class Codes: G01C 21/3617 : using user history, behaviour, conditions or preferences, e.g. predicted or inferred from previous use or current movement
• CPC Class Codes: G01S 13/865 : Combination of radar systems with lidar systems
• CPC Class Codes: G01S 13/867 : Combination of radar systems with cameras
• CPC Class Codes: G01S 17/86 : Combinations of lidar systems with systems other than lidar, radar or sonar, e.g. with direction finders
• CPC Class Codes: G01S 17/931 : of land vehicles
• CPC Class Codes: G05B 2219/2637 : Vehicle, car, auto, wheelchair
• CPC Class Codes: G05D 1/0055 : with safety arrangements
• CPC Class Codes: G05D 1/0088 : characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours using knowledge based models G06N5/00
• CPC Class Codes: G05D 1/021 : specially adapted to land vehicles
• CPC Class Codes: G05D 1/0212 : with means for defining a desired trajectory involving a plurality of land vehicles G05D1/0287
• CPC Class Codes: G05D 1/0214 : in accordance with safety or protection criteria, e.g. avoiding hazardous areas monitoring the location of vehicles within a certain area, e.g. forbidden or allowed areas, in traffic control systems for road vehicles G08G1/13
• CPC Class Codes: G05D 1/024 : in combination with a laser lasers per se H01S
• CPC Class Codes: G05D 1/0246 : using a video camera in combination with image processing means
• CPC Class Codes: G05D 1/0257 : using a radar radar systems designed for anti-collision purposes between land vehicles or between land vehicle and fixed obstacles G01S13/931
• CPC Class Codes: G05D 1/0274 : using mapping information stored in a memory device navigation using map-matching G01C21/30
• CPC Class Codes: G05D 1/0276 : using signals provided by a source external to the vehicle involving a plurality of vehicles G05D1/0287; automatically controlling vehicle speed responsive to externally generated signals B60K31/0058
• CPC Class Codes: G05D 1/0278 : using satellite positioning signals, e.g. GPS
• CPC Class Codes: G06Q 10/02 : Reservations, e.g. for tickets, services or events
• CPC Class Codes: G06Q 30/02 : Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0207 : Discounts or incentives, e.g. coupons or rebates
• CPC Class Codes: G06Q 40/08 : Insurance
• CPC Class Codes: G06T 2207/10004 : Still image; Photographic image
• CPC Class Codes: G06T 2207/10028 : Range image; Depth image; 3D point clouds
• CPC Class Codes: G06T 2207/30236 : Traffic on road, railway or crossing
• CPC Class Codes: G06T 2207/30252 : Vehicle exterior; Vicinity of vehicle
• CPC Class Codes: G06T 2207/30261 : Obstacle
• CPC Class Codes: G06T 7/20 : Analysis of motion motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51
• CPC Class Codes: G06T 7/223 : using block-matching
• CPC Class Codes: G06T 7/231 : using full search
• CPC Class Codes: G06T 7/521 : from laser ranging, e.g. using interferometry; from the projection of structured light
• CPC Class Codes: G06T 7/74 : involving reference images or patches
• CPC Class Codes: G06V 10/255 : Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
• 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 20/588 : Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
• CPC Class Codes: G06V 20/64 : Three-dimensional objects
• CPC Class Codes: G07C 9/00563 : using personal physical data of the operator, e.g. finger prints, retinal images, voicepatterns

Abstract

A method and apparatus are provided for optimizing one or more object detection parameters used by an autonomous vehicle to detect objects in images. The autonomous vehicle may capture the images using one or more sensors. The autonomous vehicle may then determine object labels and their corresponding object label parameters for the detected objects. The captured images and the object label parameters may be communicated to an object identification server. The object identification server may request that one or more reviewers identify objects in the captured images. The object identification server may then compare the identification of objects by reviewers with the identification of objects by the autonomous vehicle. Depending on the results of the comparison, the object identification server may recommend or perform the optimization of one or more of the object detection parameters.

Claims

1. A method for optimizing object detection performed by autonomous vehicles, the method comprising:

receiving, by one or more computing devices having one or more processors, sensor data captured by a sensor of an autonomous vehicle, wherein the sensor data includes one or more first object labels applied by the autonomous vehicle, each of the one or more first object labels corresponding to an object of the sensor data;

receiving, by the one or more computing devices, one or more second object labels for the sensor data;

determining, by the one or more computing devices, a difference between a first value indicating a quantity of the one or more first object labels applied by the autonomous vehicle and a second value indicating a quantity of the one or more second object labels;

determining, by the one or more computing devices, whether the autonomous vehicle has met a predetermined missed object threshold corresponding to a maximum allowable quantity of object labels not applied by the autonomous vehicle by comparing the difference to the threshold; and

when the autonomous vehicle has met the threshold, generating a recommendation for an adjustment of at least one object detection parameter of at least one of a plurality of sensors in order to optimize object detection.

2. The method of claim 1, wherein the one or more second object labels include labels generated at least in part by one or more processors of a second computing device.

3. The method of claim 1, wherein the one or more second object labels include labels generated at least in part by a human operator.

4. The method of claim 1, wherein the threshold is an absolute number of objects.

5. The method of claim 1, wherein the threshold is a percentage of the one or more second object labels.

6. The method of claim 1, further comprising displaying on a display device the recommendation.

7. The method of claim 1, further comprising:

determining an accuracy value of the one or more first object labels by comparing label types of the one or more first object labels and label types of the one or more second object labels indicate, and

wherein the recommendation is further based on the accuracy value.

8. The method of claim 7, further comprising:

determining whether the accuracy value meets an object label threshold, and

wherein the recommendation is generated further based on whether the accuracy value meets an object label threshold.

9. The method of claim 7, further comprising:

for each given first object label of the one or more second object labels, determining an object label ratio based on (1) an intersection of a given second object label and the given first object label and (2) an area of a union of the given first object label and the given second object label, and

wherein determining the accuracy value is further based on any determined object label ratios.

10. The method of claim 9, further comprising:

using any determined object label ratios to determine a mean object label ratio value, and

wherein determining the accuracy value is further based on the mean object label ratio value.

11. A system for optimizing object detection performed by autonomous vehicles, the system comprising one or more computing devices having one or more processors configured to:

receive sensor data captured by an autonomous vehicle, wherein the sensor data includes one or more first object labels applied by the autonomous vehicle, each of the one or more first object labels corresponding to an object of the sensor data;

receive one or more second object labels for the sensor data;

determine a difference between a first value indicating a quantity of the one or more first object labels applied by the autonomous vehicle and a second value indicating a quantity of the one or more second object labels;

determine whether the autonomous vehicle has met a predetermined missed object threshold corresponding to a maximum allowable quantity of object labels not applied by the autonomous vehicle by comparing the difference to the threshold; and

when the autonomous vehicle has met the threshold, generate a recommendation for an adjustment of at least one object detection parameter of at least one of a plurality of sensors in order to optimize object detection.

12. The system of claim 11, wherein the one or more second object labels include labels generated at least in part by one or more processors of a second computing device.

13. The system of claim 11, wherein the one or more second object labels include labels generated at least in part by a human operator.

14. The system of claim 11, further comprising:

determining an accuracy value of the one or more first object labels by comparing label types of the one or more first object labels and label types of the one or more second object labels indicate, and

wherein the recommendation is further based on the accuracy value.

15. The system of claim 14, further comprising:

determining whether the accuracy value meets an object label threshold, and

wherein the recommendation is generated further based on whether the accuracy value meets an object label threshold.

17. The system of claim 14, wherein the one or more processors are further configured to:

for each given first object label of the one or more second object labels, determine an object label ratio based on (1) an intersection of a given second object label and the given first object label and (2) an area of a union of the given first object label and the given second object label, and

determine the accuracy value further based on any determined object label ratios.

18. The system of claim 17, wherein the one or more processors are further configured to:

use any determined object label ratios to determine a mean object label ratio value, and

determine the accuracy value further based on the mean object label ratio value.

16. A non-transitory, tangible recording medium on which instructions are stored, the instructions, when executed by one or more processors of one or more computing devices, cause the one or more processors to perform a method for optimizing object detection performed by autonomous vehicles, the method comprising:

receiving sensor data captured by an autonomous vehicle, wherein the sensor data includes one or more first object labels applied by the autonomous vehicle, each of the one or more first object labels corresponding to an object of the sensor data;

receiving one or more second object labels for the sensor data;

determining a difference between a first value indicating a quantity of the one or more first object labels applied by the autonomous vehicle and a second value indicating a quantity of the one or more second object labels;

determining whether the autonomous vehicle has met a predetermined missed object threshold corresponding to a maximum allowable quantity of object labels not applied by the autonomous vehicle by comparing the difference to the threshold; and

when the autonomous vehicle has met the threshold, generating a recommendation for an adjustment of at least one object detection parameter of at least one of a plurality of sensors in order to optimize object detection.

19. The medium of claim 16, wherein the method further comprises:

determining an accuracy value of the one or more first object labels by comparing label types of the one or more first object labels and label types of the one or more second object labels indicate, and

wherein the recommendation is further based on the accuracy value.

20. The medium of claim 19, wherein the method further comprises:

determining whether the accuracy value meets an object label threshold, and

wherein the recommendation is generated further based on whether the accuracy value meets an object label threshold.