METHOD FOR EVALUATING AN OBJECT RECOGNITION DEVICE OF A MOTOR VEHICLE

• US 20130016216A1
• Filed: 07/03/2012
• Published: 01/17/2013
• Est. Priority Date: 07/15/2011
• Status: Active Grant

• Current Assignee: Audi AG (Volkswagen AG)
• Original Assignee: Audi AG (Volkswagen AG)
• Inventors: Nentwig, Mirko
• Granted Patent: US 9,286,525 B2
• Time in Patent Office: Days
• US Class Current: 348/148
• 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

Abstract

For testing an object recognition device for a motor vehicle at reasonable costs for different routes, image data for testing the object recognition device may be generated with a camera simulation device. Because the image data of a camera simulation device are artificially generated, it must be made certain that they have a realistic effect on the object recognition device. Reference image data are generated with a camera and simulation image data are generated with the camera simulation device for at least one route. The simulation image data and the reference image data are compared with each other based on at least two comparison measures. A value which is independent of the object recognition device to be tested can be determined for each of the comparison measures. It is then checked if the totality of the generated comparison values satisfies a predetermined validation criterion.

Claims

1. A method for evaluating with a camera simulation device operability of an object recognition device of a motor vehicle, comprising the steps of:

providing reference image data generated with a camera for at least one route;

generating with the camera simulation device simulation image data for the at least one route;

comparing the simulation image data with the reference image data based on at least two different predetermined comparison measures;

determining a value for each of the at least two comparison measures independent of processing of the data by the object recognition device;

forming at least one comparison value for each comparison measure;

checking if a totality of the formed comparison measures satisfies a predetermined validation criterion; and

when the predetermined validation criterion is satisfied, providing the camera simulation device for use in evaluating the operability of the object recognition device.

2. The method of claim 1, wherein one of the comparison measures describes a difference of color intensity data, brightness data, contrast data or edge contour direction data.

3. The method of claim 1, wherein the simulation image data are compared with the reference image data for individual, mutually corresponding image regions of two images, with one of the images being represented by the simulation image data and the other image by the reference image data, and wherein a comparison value is formed for each image region and for each comparison measure.

4. The method of claim 1, further comprising the steps of:

computing for at least a portion of the simulation image data a first statistical average value and/or a first statistical variance value,

computing for a corresponding portion of the reference image data a second statistical average value and/or a second statistical variance value, and

forming one of the comparison values depending on the computed statistical values.

5. The method of claim 1, further comprising the steps of:

computing a first histogram for at least a portion of the simulation image data,

computing a second histogram for a corresponding portion of the reference image data, and

forming one of the comparison values depending on the computed histograms.

6. The method of claim 1, wherein the step of checking if the predetermined validation criterion is satisfied, further comprises linking comparison values formed for different comparison measures with each other to form an overall comparison value.

7. The method of claim 6, wherein before linking, at least one of the comparison values is scaled with a predetermined weighting factor.

8. The method of claim 1, further comprising iteratively adapting operation of the camera simulation device for use in the evaluation by

defining the operation of the camera simulation device with a configuration data set,

determining for the configuration data set if the predetermined validation criterion is satisfied, and

changing the configuration data set commensurate with the validation result.

9. The method of claim 1, further comprising the steps of:

graphically representing at least a portion of the comparison values on a display device, and

associating the displayed comparison values with image regions corresponding to simulation and reference image data that form a basis for the displayed comparison values.

10. The method of claim 1, wherein the simulation image data are compared with the reference image data based on at least three different predetermined comparison measures.

11. The method of claim 1, wherein the validation criterion is defined based on validation image data which are known to result in meaningful evaluation results when evaluating the operability of the object recognition device.

12. The method of claim 1, and further comprising the steps of:

defining a value range, and

defining a validation criterion for the defined value range,

wherein each comparison value from a predetermined set or from a link of the comparison values from the predetermined set is located inside the value range.

13. A device for testing suitability of a camera simulation device for use in evaluating operability of an object recognition device of a motor vehicle, comprising:

at least two comparison devices configured to compare simulation image data of the camera simulation device with reference image data of a camera based on a predetermined comparison measure, wherein a value is determined independent of processing of the simulation image data and the reference image by the object recognition device; and

an evaluation device operatively coupled with the at least two comparison devices for checking if a totality of comparison values generated by the comparison devices for the simulation image data satisfies a validation criterion.