Method and device for classifying an object in an image

1. A driver assistance system, comprising:

at least one camera for generating images of a surrounding of a vehicle;

a data processing unit is configured to:

define an image area located within the object in the image,

decompose the image area into an array of subareas,

define an array of aggregated pixel values by calculating, for each of the subareas, an aggregated pixel value of the respective subarea,

calculate a gradient array depending on differences between the aggregated pixel values of adjacent subareas,

analyze the gradient array, and

identify depending on a result of the analyzing the gradient array, the object as belonging to a class of a predefined set of classes; and

a driver information unit coupled to the data processing unit and providing information on the objects classified as traffic signs and/or speed limit signs to a driver of the vehicle,

wherein the aggregated pixel values of the subareas are obtained by summing up pixel values of pixels contained in the respective subarea.

2. The system of claim 1 , wherein the object in the image is a traffic sign or a candidate for a traffic sign and wherein the predefined set of classes includes a set of classes of traffic signs.

3. The system of claim 2 , wherein the set of classes of traffic signs includes at least one class of speed limit signs, each of the speed limit signs of this class displaying a number with one-digit or two-digit or three-digit number.

4. The system of claim 3 , wherein the data processing unit is configured to analyze the object identified as belonging to the at least one class of speed limit signs for identifying the number displayed on this particular speed limit sign.

5. The system of claim 1 , wherein the image area is or comprises a horizontal band.

6. The system of claim 1 , wherein the subareas of the image area are vertical strips.

7. The system of claim 1 , wherein the analyzing the gradient array includes comparing gradient values of the gradient array with at least one threshold value, defining a feature array by mapping the gradient values onto a smaller set of values depending on whether the respective value of the gradient array is larger or smaller than the at least one threshold value, and analyzing the feature array.

8. The system of claim 1 , wherein the data processing unit is configured to compare features of the gradient array with a predefined set of features.

9. The system of claim 8 , wherein the predefined set of features is obtained from ground truth data using a training set of objects.

10. A vehicle comprising the driver assistance system of claim 1 , wherein the at least one camera is installed in the vehicle to capture the surrounding of the vehicle in front of the vehicle.

11. A method for classifying an object in an image, comprising:

generating, by at least one image capturing device, an image around a vehicle,

defining, by a data processing unit connected to the image capturing device, an image area located within the object in the image,

decomposing, by the data processing unit, the image area into an array of subareas,

defining, by the data processing unit, an array of aggregated pixel values by calculating, for each of the subareas, an aggregated pixel value of the respective subarea,

calculating, by the data processing unit, a gradient array depending on differences between the aggregated pixel values of adjacent subareas,

analyzing, by the data processing unit, the gradient array,

identifying, by the data processing unit, depending on a result of the analyzing the gradient array, the object as belonging to a class of a predefined set of classes, and

providing, by a driver information unit, information related to the object to a driver of the vehicle,

wherein the aggregated pixel values of the subareas are obtained by summing up pixel values of pixels contained in the respective subarea, and

wherein the object in the image is a traffic sign or a candidate for a traffic sign and wherein the predefined set of classes includes a set of classes of traffic signs.

12. The method of claim 11 , wherein the set of classes of traffic signs includes at least one class of speed limit signs, each of the speed limit signs of this class displaying a number with one-digit or two-digit or three-digit number.

13. The method of claim 12 , further comprising analyzing the object identified as belonging to the at least one class of speed limit signs for identifying the number displayed on this particular speed limit sign.

14. The method of claim 11 , wherein the image area is or comprises a horizontal band.

15. The method of claim 11 , wherein the subareas of the image area are vertical strips.

16. The method of claim 11 , wherein the analyzing the gradient array includes comparing gradient values of the gradient array with at least one threshold value, defining a feature array by mapping the gradient values onto a smaller set of values depending on whether the respective value of the gradient array is larger or smaller than the at least one threshold value, and analyzing the feature array.

17. The method of claim 1 , wherein the analyzing the gradient array includes comparing features of the gradient array with a predefined set of features.