Patent Application
20110157362
This application is the U.S. national phase application of PCT International Phase Application No. PCT/DE2009/000579, filed Apr. 27, 2009, which claims priority to German Patent Application No. 10 2008 022 064 7, filed May 3, 2008, the contents of such applications being incorporated by reference herein.
The invention relates to a method for exposure control for a camera in a motor vehicle and is applied, for example, in vehicles that are equipped with driver assistance functions.
An optimally illuminated image is the prerequisite for the reliable and perfect detection of objects and is important for the reliability of a driver assistance system.
US 2004/0227814 A1, which is incorporated herein by reference, shows a double exposure method for a camera system of a vehicle that synthesizes first and second images obtained using different exposure times.
DE 102006028624 A1, which is incorporated herein by reference, shows an exposure-adaptive image sensor with a control device, said control device enabling the integration time to be reduced effectively.
The synthesis of two images that were recorded at different times results in the blurring of objects that change their positions quickly.
If illumination is insufficient, a reduced integration time may result in objects or markings that are poor in contrast and, as a result, cannot be reliably detected any more.
It is therefore an object of the present invention to provide a method for exposure of a camera, said method enabling objects to be reliably detected.
According to aspects of the invention, an object of the invention is achieved by a device and a method described herein.
A method for exposure control for a camera system in a motor vehicle is provided. The camera system records a sequence of images of the surroundings of the vehicle using an optimal exposure time. The optimal exposure time is predetermined in dependence on the brightness of the surroundings or on the brightness of individual objects in the image. In addition, the exposure time may depend on a provided application, e.g. a driver assistance function.
Moreover, an intermediate image is recorded using a short exposure time, i.e. an exposure time shorter than the optimal exposure time, if the optimal exposure time is longer than a predetermined threshold value.
At least two driver assistance functions are executed on the basis of the image data of the camera system, wherein the image data recorded using the optimal exposure time is used for a first driver assistance system and the image data recorded using the short exposure time is used for a second driver assistance system.
In one exemplary embodiment of the invention, the predetermined threshold value depends on the speed of the motor vehicle. In some embodiments of the present invention a small threshold value is selected when the speed is high. For example, the interrelationship between the predetermined threshold value and the speed is inversely proportional or follows a step function.
In another exemplary embodiment of the invention, images are recorded using short and long exposure times in an alternating manner.
In another exemplary embodiment of the invention, the method is used in a camera system for traffic sign recognition and, as the case may be, for a further driver assistance function.
In the following, the invention is explained on the basis of exemplary embodiments with the driver assistance functions A, B and C and figures. Driver assistance function A is a lane detection function, for example. Driver assistance function B is a light control function, for example. Driver assistance function C is a traffic sign recognition function, for example.
The invention is best understood from the following detailed description when read in connection with the accompanying drawings. Included in the drawings is the following figures:
During the day, applications like lane detection (SpE) and traffic sign recognition (VZE) make similar demands on exposure control: moderate demands on the dynamic range of the image sensor in the sunshine and when there is hard shadow, and high demands on the dynamic range when a vehicle enters and leaves a tunnel, for example, wherein the highest possible contrast for lane detection is to be maintained.
The night represents the other extreme for exposure control. An exposure time that is as long as possible is required for lane detection. The dynamic range must be very high in order to prevent the headlights of oncoming vehicles from causing undesirable dazzle. For applications running at night, e.g. high beam assistance (FLA), exposure must be adjusted in such a manner that no objects except for oncoming vehicles cause the saturation of the image sensor. On the other hand, the exposure time for traffic sign recognition must not become too long when the speed is high, since otherwise the meaning of the traffic sign, e.g. a speed limit, cannot be clearly identified any more on account of the motion blur.
The guide value for the exposure time for avoiding the motion blur is 5 ms at a speed of 200 km/h. The conflicting demands for lane detection and traffic sign recognition at dawn/dusk and at night require a solution, which solution is explained in the following.
As a first exemplary embodiment, a method is described in which, above a predetermined speed of the motor vehicle and above a predetermined exposure time for traffic sign recognition, a second image is recorded using a short exposure time.
The exposure time for the traffic signs at night may be reduced, since the traffic signs reflect the light of the headlights at least partially and thus reflect much more light of the headlights than the asphalt of the road, for example. Even with a reduced exposure time, the traffic signs are still visible for the camera with sufficient contrast. As a rule, the refresh rate is constant for all applications, e.g. 25 Hz.
It is provided that, if an exposure time longer than a predetermined threshold value, e.g. longer than 5 ms, becomes necessary, a second image may be recorded (using a short exposure time, shorter than the predetermined threshold value, e.g. 5 ms) within the available time window (40 ms per image with a refresh rate of 25 frames per second) in order to additionally generate an image that is not distorted by motion blur. With this approach, the readout time of the image must be considerably shorter than the available time window (e.g. 15 ms with 25 frames per second), thereby enabling the first frame to be reduced to 25 ms and a second frame of 15 ms to be inserted, for example.
So the refresh rate is doubled for exposure times longer than a predetermined threshold value (e.g. 5 ms), wherein images are recorded for lane detection or high beam assistance and traffic sign recognition in an alternating manner.
A second exemplary embodiment represents a method with a simplified synchronization demand on clocking and on image evaluation. Both frames have the same length. An extended clock rate might be necessary in dependence on the respective application.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2008 022 064 7 | May 2008 | DE | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/DE09/00579 | 4/27/2009 | WO | 00 | 1/14/2011 |
