Method and apparatus for detecting a rear vehicle light

• US 9,317,757 B2
• Filed: 06/15/2010
• Issued: 04/19/2016
• Est. Priority Date: 06/15/2009
• Status: Active Grant

• Current Assignee: CARIAD SE (Volkswagen AG)
• Original Assignee: Hella KGaA Hueck & Company
• Inventors: Winter, Jan, Feiden, Dirk
• Primary Examiner(s): RAHAMAN, MOHAMMED S
• Application Number:
  US13/378,572
• Publication Number:
  US 20120176499A1
• Time in Patent Office: 2,135 Days
• Field of Search: 348/148
• US Class Current: 1/1
• CPC Class Codes: G06V 20/584 of vehicle lights or traffic lights

Abstract

A state of a rear vehicle light is detected. A camera system integrated in a vehicle generates an image of an environment in front of the vehicle. An evaluation unit connected to the camera system detects at least one vehicle contained in the image, defines an image range associated with said vehicle, and divides the image range into sub-ranges comprising at least a left sub-range and a right sub range. A segmentation of the sub-ranges into image segments is performed by way of the evaluation unit on the basis of brightness values of points contained in said sub-ranges. At least one brightest image segment is extracted from each of said sub-ranges, and the extracted image segments are checked for whether they can be associated with a particular vehicle light in an ON state.

Claims

1. A method for detecting a state of a rear vehicle light, wherein the rear vehicle light comprises a switched on turn indicator and/or switched on hazard lights, and wherein the method comprises:

receiving an image of an environment in front of a vehicle;

detecting at least one vehicle contained in the image;

defining an image region of the image associated with the at least one vehicle;

dividing the image region into part regions that include at least one left part region and at least one right part region;

segmenting the left and right part regions into image segments based on brightness values of pixels contained in the left and right part regions;

extracting at least one brightest image segment from each of the left and right part regions;

analyzing the extracted image segments as to whether they can be associated with the rear vehicle light in a switched on state;

extracting a time sequence of the at least one brightest image segment from the left part region and right part region;

analyzing a time periodicity of the at least one brightest image segment from the left part region and right part region; and

generating a signal if the rear vehicle light is in the switched on state and if the periodicity has a frequency within a predefined frequency range,

wherein the method further comprises;

comparing the extracted time sequence of the at least one brightest image segment from the left part region with the extracted image time sequence of the at least one brightest image segment from the right part region; and

generating a signal to indicate that the hazard lights are switched on if the extracted time sequence of the at least one brightest image segment from the left and right part regions are synchronized,

wherein the left and right part regions include rows of pixels, and wherein the segmenting comprises;

defining rows of pixels in the left and right part regions, wherein each pixel includes a brightness value;

sorting the rows of pixels into row segments based on the brightness values of the pixels such that each row segment includes pixels of substantially similar brightness values; and

generating image segments by pooling mutually contacting row segments of adjacent rows of pixels in the left and right part regions.

2. The method of claim 1, wherein the at least one brightest image segment is only extracted when brightness values of pixels of the at least one brightest image segment exceed a threshold brightness value.

3. The method of claim 1, wherein the segmentation step comprises:

defining limits between the image segments where brightness edges are detected.

4. The method of claim 1, wherein the detecting step comprises:

detecting at least one vehicle with image data from a lidar sensor or of a radar sensor.

5. The method of claim 1, and further comprising:

controlling at least one controllable device with the signal.

6. The method of claim 1, wherein the predefined frequency range has a frequency in the range of about 1 Hz to about 2 Hz.

7. The method of claim 1, and further comprising:

determining whether the extracted image segment from the left part region and the extracted image segment from the right part region are of the same size and are at the same level.

8. A method for detecting a state of a rear vehicle light, wherein the rear vehicle light comprises a brake light, and wherein the method comprises:

receiving an image of an environment in front of a vehicle;

detecting at least one vehicle contained in the image;

defining an image region of the image associated with the at least one vehicle;

dividing the image region into part regions that include at least one left part region, one right part region, and one middle part region;

segmenting the left, right, and middle part regions into image segments based on brightness values of pixels contained in the left, right, and middle part regions, respectively;

extracting at least one brightest image segment from each of the left, right, and middle part regions;

analyzing the extracted image segments as to whether they can be associated with the rear vehicle light in a switched on state; and

generating a signal if the rear vehicle light is in the switched on state and if (a) the at least one brightest image segment from the left part region and the at least one brightest image segment from the right part region are of the same size and are at the same level and (b) the at least one brightest image segment from the middle part region is located higher and centrally between the at least one brightest image segments from the left and right part regions,

wherein the left and right part regions include rows of pixels, and wherein the segmenting comprises;

defining rows of pixels in the left and right part regions, wherein each pixel includes a brightness value;

sorting the rows of pixels into row segments based on the brightness values of the pixels such that each row segment includes pixels of substantially similar brightness values; and

generating image segments by pooling mutually contacting row segments of adjacent rows of pixels in the left and right part regions.

9. The method of claim 8, wherein one brightest image segment is extracted in the middle part segment and one or two or three brightest image segments are extracted in each of the left and right part regions.

10. A system for detecting a state of a rear vehicle light, wherein the rear vehicle light comprises a switched on turn indicator and/or switched on hazard lights, and wherein the system comprises:

a camera system integrated in a vehicle configured to generate an image of an environment in front of the vehicle; and

an evaluation unit connected to the camera system, the evaluation unit configured to;

detect at least one vehicle contained in the image;

define an image region of the image associated with this vehicle;

divide the image region into part regions that include at least one left part region and one right part region;

segment the left and right part regions into image segments based on brightness values of pixels contained in the left and right part regions;

extract at least one brightest image segment from each of the left and right part regions;

analyze the extracted image segments as to whether they can be associated with the rear vehicle light in a switched on state;

extract a time sequence of the at least one brightest image segment from the left part region and right part region;

analyze a time periodicity of the at least one brightest image segment from the left part region and right part region; and

generate a signal if the rear vehicle light is in the switched on state and if the periodicity has a frequency within a predefined frequency range,

wherein the evaluation unit is further configured to;

compare the extracted time sequence of the at least one brightest image segment from the left part region with the extracted image time sequence of the at least one brightest image segment from the right part region; and

generate a signal to indicate that the hazard lights are switched on if the extracted time sequence of the at least one brightest image segment from the left and right part regions are synchronized,

wherein, to segment the left and right part regions into image segments, the evaluation unit is configured to;

define rows of pixels in the left and right part regions, wherein each pixel includes a brightness value;

sort the rows of pixels into row segments based on the brightness values of the pixels such that each row segment includes pixels of substantially similar brightness values; and

generate image segments by pooling mutually contacting row segments of adjacent rows of pixels in the left and right part regions.

11. The system of claim 10, and further comprising:

a lidar sensor and/or a radar sensor connected to the evaluation unit and configured to provide image data to the evaluation unit for the detection of the at least one vehicle contained in the image.

12. The system of claim 10, wherein the evaluation unit is connected to at least one device controllable by the generated signal.

13. The system of claim 10, wherein the camera system comprises a grayscale camera to generate grayscale images.

14. A system for detecting a state of a rear vehicle light, wherein the rear vehicle light comprises a brake light, and wherein the system comprises:

a camera system integrated in a vehicle configured to generate an image of an environment in front of the vehicle; and

an evaluation unit connected to the camera system, wherein the evaluation unit is configured to;

detect at least one vehicle contained in the image;

define an image region of the image associated with this vehicle;

divide the image region into part regions that include at least one left part region, one right part region, and one middle part region;

segment the left, right, and middle part regions into image segments based on brightness values of pixels contained in the left, right, and middle part regions, respectively;

extract at least one brightest image segment from each of the left, right, and middle part regions;

analyze the extracted image segments as to whether they can be associated with the rear vehicle light in a switched on state; and

generate a signal if the rear vehicle light is in the switched on state and if (a) the at least one brightest image segment from the left part region and the at least one brightest image segment from the right part region are of the same size and are at the same level and (b) the at least one brightest image segment from the middle part region is located higher and centrally between the at least one brightest image segments from the left and right part regions,

wherein, to segment the left and right part regions into image segments, the evaluation unit is configured to;

define rows of pixels in the left and right part regions, wherein each pixel includes a brightness value;

sort the rows of pixels into row segments based on the brightness values of the pixels such that each row segment includes pixels of substantially similar brightness values; and

generate image segments by pooling mutually contacting row segments of adjacent rows of pixels in the left and right part regions.