WO2020094551A1

WO2020094551A1 - Method for operating a fisheye camera by use of a first compression algorithm and a second compression algorithm, electronic computing device, fisheye camera, and driver assistance system

Info

Publication number: WO2020094551A1
Application number: PCT/EP2019/080061
Authority: WO
Inventors: Fabian BURGER
Original assignee: Valeo Schalter Und Sensoren Gmbh
Priority date: 2018-11-05
Filing date: 2019-11-04
Publication date: 2020-05-14
Also published as: DE102018127465A1

Abstract

The invention relates to a method for operating a fisheye camera (4) for a driver assistance system (3) of a motor vehicle (1), in which a fisheye image (12) is captured by means of the fisheye camera and the fisheye image (12) is analysed according to at least a first image region (13) and a second image region (14) for the driver assistance system (3) by means of an electronic computing device (11), the first image region (13) being predefined by a first distance range (15) from the fisheye camera (4) and the second image region (14) being predefined by a second distance range (16), different to the first distance range (15), from the fisheye camera (4), and the first image region (13) being compressed with a first compression algorithm (17) and the second image region (14) being compressed with a second compression algorithm (18), different to the first compression algorithm, and an image (19', 19'', 19''') compressed according to the compressed first image region (13) and the compressed second image region (14) being generated by means of the electronic computing device (11) and being transferred to the driver assistance system (3) for analysis.

Description

Method for operating a fisheye camera by using a first compression algorithm and a second compression algorithm, electronic computing device, fisheye camera and driver assistance system

The invention relates to a method for operating a fisheye camera for a driver assistance system of a motor vehicle, in which a fisheye image is captured by means of the fisheye camera and the fisheye image as a function of at least a first image area of the fisheye image and a second image area of the

Fish eye image for the driver assistance system using an electronic

Computing device of the fisheye camera is evaluated. The invention further relates to an electronic computing device, a fisheye camera and a

Driver assistance system. It is known from motor vehicle construction that to record the surroundings of the

Motor vehicle cameras are used. These cameras are, in particular, due to the increased field of view (FOV), in particular as

Fisheye cameras trained. The fisheye cameras are of particular importance in the area of at least partially autonomous, in particular fully autonomous, driving. For example, the fisheye camera can be used to carry out an optical flow, a motor vehicle or pedestrian detection or a lane detection. Editing the corresponding

Fisheye image of the fisheye camera can, for example, by means of a

Computer vision algorithm and / or by means of learning processes of deep neural networks (deep learning), for example by means of a convoluted neural network

(convolutional neural network - CNN). These evaluation algorithms are limited in particular on specially embedded platforms or on so-called digital signal processors (digital signal processors - DSP)

Computing capacity carried out.

A plurality of such cameras is often arranged on the motor vehicle, for example in a front region, in a rear region and in the side region of the motor vehicle, an evaluation of this large number of carts then having to be carried out in particular simultaneously. Especially due to the limited

Computing capacity, this can often lead to considerable time delays when evaluating the fisheye images. However, there are also disadvantages, particularly when using the fisheye camera. For example, the fish-eye image shows the Fisheye distortion. The distortion is particularly noticeable in vertical lines. Furthermore, nearby objects are shown enlarged. These fisheye images are corrected and straightened in particular, so that, for example, the computer vision algorithm can easily evaluate the corrected image. However, these corrected images also continue to show certain distortions, particularly in the case of horizontal lines and in close-up areas.

In order to save computing time, it is known that normally the entire image of the fisheye camera is not evaluated. Different strategies are known from the prior art, which in particular evaluate a so-called region of interest (ROI) of the image accordingly. In the prior art, only a partial area of the fisheye image, in which the important parts for further evaluation are located in particular, is processed and the other image areas are not further evaluated, this being the case

is particularly dependent on the task of the driver assistance system. In particular, the computer vision algorithm is then only used in the regions of interest.

Evaluating regions of interest also involves errors because

For example, static regions of interest cannot recognize objects that are outside the boundary of the region of interest, for example. Further evaluations using regions of interest in particular show an increased computing effort, since it is possible, for example, that a partial area of the image occurs in a plurality of regions of interest and is therefore also evaluated several times.

The object of the present invention is a method, an electronic

To create computing device, a fisheye camera and a driver assistance system, by means of which computing capacity can be saved.

This task is accomplished by a method, an electronic computing device, a fisheye camera and a driver assistance system according to the independent

Claims resolved.

One aspect of the invention relates to a method for operating a fisheye camera for a driver assistance system of a motor vehicle, in which the

Fisheye camera a fisheye image is captured and the fisheye image in Dependency of at least a first image area of the fisheye image and a second image area of the fisheye image for the driver assistance system is evaluated by means of an electronic computing device of the fisheye camera.

It is provided that the first image area is specified by a first distance area from the fisheye camera and the second image area is specified by a second distance area from the fisheye camera that differs from the first distance area, and the first image area with a first compression algorithm and the second image area with a second compression algorithm that is different from the first compression algorithm be compressed and an image compressed as a function of the compressed first image area and the compressed second image area is generated by means of the electronic computing device and sent to the

Driver assistance system for evaluation is transmitted.

In particular, the fisheye image is thus compressed as a function of the respective image area of the fisheye image, the image area being in particular dependent on the distance. For example, it can be provided that a close-up area of the fisheye camera, which due to the arrangement on the motor vehicle is an area directly in front of the motor vehicle, is compressed correspondingly highly, since in particular little information is available here, and therefore a small one

Loss of information during compression takes place. In particular, an area further away from the fisheye camera can be identified as an image area and this in turn can be correspondingly less compressed, in particular decompressed, that is to say a negative compression, so that a higher image area

Density of information can be generated. In particular, these two compressed image areas are then combined into the common compressed image and transferred to the driver assistance system for transmission.

The driver assistance system can then use the computer vision algorithm, for example, to process the compressed image, which can save computing time. In particular, the computing time is saved without significant

Having to accept loss of information. This enables a more efficient use of the electronic computing device. In particular, this enables a faster evaluation of the fish eye image. Furthermore, it is possible that, for example, processors are used in the electronic computing device, which may have a lower computing power, as a result of which for example space or costs related to the electronic

Computing device can be saved.

In particular, it can be provided that the first compression algorithm and / or the second compression algorithm are used to compress the first image area and / or the second image area with respect to the data volume of the

Fisheye picture takes place. In other words, the image can be compressed in particular with respect to, for example, the respective pixels in the first image area and / or in the second image area. A smaller amount of data is thus provided in the compressed image, which can then be transmitted to the driver assistance system.

Furthermore, it can be provided that the visually visible image can also be compressed in accordance with its size, as a result of which the amount of data which is transmitted to the driver assistance system is in turn compressed accordingly.

It is preferably possible that compression is carried out both in the vertical direction and in the horizontal direction. For example, it can be provided that a vanishing point of the fisheye image is used and depending on a relative position of the first image area and / or the second image area

Vanishing point the respective compression is performed. In particular, it can be provided that no compression is carried out in the vanishing point, however, the further, viewed in both the vertical and horizontal directions, an image pixel from the

Vanishing point in the fisheye image removed, the greater a corresponding compression is carried out. In particular, it can be provided that the fisheye image is divided into a plurality of image areas, in particular into individual pixels, and the plurality of different image areas are correspondingly different

Compression algorithms is compressed. This makes it reliable and

Information-preserving compression of the fisheye image enables.

In particular, a flowing compression can take place between the first image area and the second image area. In other words, essentially the same compression takes place at least at a common border area of the first image area and the second image area. In particular, linear compression is thus recorded in neighboring image areas, which is preferred.

In particular, the compressed image still has the same objects in the

Surround the fisheye camera on how the fisheye picture taken while the compressed image is made significantly smaller. Furthermore, the compressed image has a difference in the object size, in particular objects that are further away appear larger than objects that appear close, in particular the objects that are close are shown smaller, which is particularly for

Object detection can be advantageous.

According to an advantageous embodiment, in addition to generating the compressed image, fisheye correction is performed using the electronic one

Computing device performed. In other words, in addition to the compression of the first image area and the second image area to the compressed image, the fisheye image is also corrected. In particular, this is already carried out by means of the electronic computing device of the fisheye camera, so that the driver assistance system can be provided with a compressed and corrected image. This has the particular advantage that the driver assistance system can already evaluate the compressed and corrected image, so that a

computing capacity-saving and reliable evaluation of the fisheye image can be carried out by means of the driver assistance system.

It is also advantageous if the fisheye correction is carried out before the generation of the compressed image. This enables the image to be compressed only after the fisheye correction has been carried out. This enables an improved compression of the image.

It has also proven to be advantageous if the first distance range is

The near range of the fisheye camera is specified and the second distance range is specified as the far range of the fisheye camera. In other words, the first distance area can be located directly in front of the fisheye camera and the second

The distance range can be further away from the fisheye camera.

For example, the first distance range can be defined by a distance of one meter to two meters, in particular one meter to three meters, in particular one meter to five meters. The second distance range can be defined, for example, from two meters to 100 meters, in particular from three meters to 100 meters, in particular from five meters to 100 meters. The enumeration of the definition of

Distance ranges are purely exemplary and are in no way to be regarded as conclusive.

In particular, by specifying the distance ranges, the first compression algorithm and / or the second compression algorithm can be applied to the

Close range and / or applied to the far range. Especially since Near area of the fisheye camera contains less environmental information than in the far area of the fisheye camera, it can thus be provided, for example, that the near area is compressed more than the far area. In particular, there is hardly any loss of information about the surroundings, but the amount of data within the transmitted compressed image can be significantly reduced, so that the driver assistance system is confronted with fewer amounts of data for evaluation.

A simple, yet reliable compression of the fisheye image can thus take place.

Furthermore, it has proven to be advantageous if a compression of the near area is greater than a compression of the far area. In other words, there is greater compression in the near area than in the far area. In particular, for example, the far area can include the vanishing point of the fisheye camera.

In particular, no compression takes place in the vanishing point, while the compression increases with a distance from the vanishing point. In other words, it is particularly provided that within the close range and / or

Long range non-linear compression is performed. In particular, this prevents information losses from taking place in the long range, while the information density in the close range increases, for example, so that the surroundings can still be reliably detected and, for example, object detection can be carried out using the driver assistance system.

The close-up and far-away areas preferably have a

common border on, so that a smooth transition from long range to

Close range is realized. In particular, the image only has the far and near areas.

In a further advantageous embodiment, the first image area and / or the second are made by means of a non-linear hyperbolic projection algorithm

Image area compressed. In other words, it is the first

Compression algorithm around a non-linear hyperbolic

Projection algorithm. In particular, it can also be the second

Compression algorithm is a non-linear hyperbolic projection algorithm. In other words, it is provided that the compression is thus carried out non-linearly in the first image area and / or in the second image area. By means of the non-linear hyperbolic projection algorithm, the respective strength of the compression, in particular pixel by pixel, can be realized and reproduced. By means of the non-linear hyperbolic projection algorithm can be a reliable one

Compression algorithm are provided, by means of which computing effort can be reduced and the compressed image can be evaluated by means of the driver assistance system.

It has also proven to be advantageous if at least a third image area of the fisheye image by means of at least one third compression algorithm, which is different from the first compression algorithm and the second

Compression algorithm is compressed and one depending on the

compressed first image area, the compressed second image area and the at least compressed third image area compressed image by means of the

electronic computing device is generated and transmitted to the driver assistance system for evaluation. For example, the third image area can be a sky area in the captured fish-eye image. In particular, since little information can be recorded in this sky area for the detection of the surroundings or for object detection, it can be provided, for example, that the third image area is compressed very strongly, since the probability is high here that there is no essential environmental information for evaluation with the

Driver assistance system is lost. This enables it to continue

Computing capacity can be saved and still reliable object detection in the relevant image areas or reliable evaluation can be carried out using the driver assistance system.

It has also proven to be advantageous if, depending on a function of the driver assistance system, a parameter of the first compression algorithm for compressing the first image area and / or a parameter of the second

Compression algorithm for compressing the second image area can be selected by means of the electronic computing device. If, for example, the function of the driver assistance system is to recognize pedestrians, an area in front of the motor vehicle, in particular in the vicinity of the motor vehicle, is of particular interest, so that the compression algorithm is then selected such that only a small loss of information takes place in this area. If, for example, the function of the driver assistance system is such that motor vehicles are to be recognized by the fisheye image, this takes place in particular in the far area, so that compression of the far area is prevented

or a decompression of the far area can be carried out and the near area can be compressed more, for example. So in particular the parameter of the first compression algorithm and / or the second

Compression algorithm is a parameter for the strength of the compression. Depending on the driver assistance system, the fisheye image can be compressed accordingly and for evaluation for the driver assistance system

be provided in a time-saving manner.

It has also proven to be advantageous if at least one extrinsic and / or intrinsic parameter of the fisheye camera is stored in a look-up table of the electronic computing device and fisheye correction is carried out as a function of the stored extrinsic and / or the stored intrinsic parameter. The lookup table which

can also be referred to in particular as a look-up table, in particular has a large number of extrinsic and / or intrinsic parameters of the fisheye camera, as a result of which a reliable correction of the image with respect to the fisheye distortion can be carried out. In particular, the fisheye correction already takes place within the electronic computing device, so that time can still be saved by the driver assistance system when evaluating the received image. In particular, fisheye correction is highly non-linear. In particular, the

Look-up table for each pixel of the uncorrected image a corresponding pixel of the corrected image. The look-up table is specified in particular, this being possible, for example, by means of corresponding reference measurements. This makes it possible that only the memory of the electronic computing device has to be used during the fisheye correction in order to carry out the corresponding fisheye correction. In particular, this saves computing time during the fisheye correction.

In particular, it can be provided that the first compression algorithm and / or the second compression algorithm are likewise stored in the look-up table with corresponding parameters, and the compression can also be carried out by means of the look-up table on the basis of the stored data in the look-up table. This makes it possible that no additional runtime or computing capacity is required to carry out the compression of the image.

Furthermore, it has proven to be advantageous if the first image area and / or the second image area are determined as a function of a detection frequency of an occurrence of a specific object. For example, it can be provided that as certain object a pedestrian is specified. In particular are

Pedestrians, which are important for the driver assistance system, in the vicinity of the fisheye camera. In particular, the distance area for the first image area and / or for the second image area can thereby be made dependent on this object. If, for example, the object is specified as a motor vehicle, the respective distance range can be enlarged, since in particular the detection of motor vehicles in the immediate vicinity of the fisheye camera is of less importance than the detection of motor vehicles in a distant environment from the fisheye camera. In particular, the definition of the first image area and in particular the associated distance area and / or the second image area and in particular the associated second distance area is thus defined by the respective object to be detected or by the function of the driver assistance system. In particular, the frequency of detection

For example, it is provided that the occurrence of a pedestrian in the near range is greater than in the far range, while, for example, the occurrence of the motor vehicle in the far range is higher than in the close range. This is a situation dependent

Compress the image.

It is also advantageous if a parameter of the first compression algorithm for compressing the first image area and / or a parameter of the second

Compression algorithm for compressing the second image area by specifying an image size and / or a processing time of the compressed image for the driver assistance system can be specified by the driver assistance system. In other words, it can be provided, for example, that the driver assistance system only requires an image size of a predetermined number of pixels for processing, in which case the compression of the first image area and the second image area can then be carried out as a function of this image size. Furthermore, it can be provided, for example, that the driver assistance system only evaluates a predetermined processing time for evaluating the transmitted image and, depending on this, the corresponding data size is again calculated and the image is compressed accordingly in order to be able to achieve this processing time.

This makes it possible for the compressed image to be adapted reliably and as a function of a desired processing time and / or a predetermined image size of the driver assistance system.

In a further advantageous embodiment, the first

Compression algorithm and the second compression algorithm a compression of the fisheye image between 20% and 70%. In particular, the compression can be carried out within this range of 20% and 70% of the original image size, the image size corresponding in particular to the original amount of data. This enables that depending on the one you want

Information density and / or the desired processing time, a corresponding compression of the image can be carried out. This has the advantage, for example, if an image with little is only required by the driver assistance system

Surrounding information, a compression of 70% can be selected, so that the driver assistance system in particular can quickly evaluate the image accordingly. If, for example, the driver assistance system wishes that a high density of surrounding information is maintained, the image can only be compressed by 20%, so that the compressed image still has a high information density and can nevertheless be processed faster than the uncompressed image. This can depend on the situation and the function of the

Driver assistance system can be used to compress the image.

Another aspect of the invention relates to an electronic computing device which is designed to carry out the method according to the preceding aspect. In particular, the method is carried out by means of the electronic computing device.

A still further aspect of the invention relates to a computer program product with program code means which are stored in a computer-readable medium in order to carry out the method for operating a fisheye camera according to the preceding aspect when the computer program product is processed on a processor of the electronic computing device.

Yet another aspect of the invention relates to a fisheye camera with an electronic computing device according to the previous aspect.

Yet another aspect of the invention relates to a driver assistance system with a fisheye camera. In the case of partially autonomous operation or fully autonomous operation of the motor vehicle, the driver assistance system can also be referred to as an electronic vehicle guidance system. Yet another aspect of the invention relates to a motor vehicle with a

Driver assistance system according to the previous aspect. The motor vehicle is designed in particular as a passenger car.

Advantageous embodiments of the method are advantageous

To see designs of the electronic computing device, the fisheye camera and the driver assistance system. For this purpose, the electronic computing device, the fisheye camera and the driver assistance system have physical features which enable the method or an advantageous embodiment thereof to be carried out.

Further features of the invention result from the claims, the figures and the description of the figures. The features and combinations of features mentioned above in the description and the features and combinations of features mentioned below in the description of the figures and / or shown alone in the figures can be used not only in the combination indicated in each case, but also in other combinations, without the scope of the invention leave. There are thus also embodiments of the invention to be regarded as encompassed and disclosed, which are not explicitly shown and explained in the figures, but can be derived from the explanations given and can be generated by separate combinations of features. There are also designs and combinations of features to be regarded as disclosed, which therefore do not have all the features of an originally formulated independent claim. Furthermore, versions and combinations of features, in particular those explained above, are to be regarded as disclosed, which go beyond or differ from the combinations of features set out in the references of the claims.

In the following, embodiments of the invention are made more schematic

Drawings explained in more detail:

Show:

Fig. 1 is a schematic plan view of a motor vehicle with a

Embodiment of a driver assistance system;

Fig. 2 is a schematic view of a captured image

Embodiment of a fisheye camera; Fig. 3 is a schematic block diagram of an embodiment of the

method according to the invention;

4 shows a schematic view of an embodiment of the method;

5 shows a further schematic view of an embodiment of the method; 6 shows a further schematic view of a recording of an embodiment of a fisheye camera; and

7 shows a schematic diagram to illustrate the method.

Identical or functionally identical components are provided with the same reference symbols in the figures.

1 shows a motor vehicle 1 in a top view. In the present case it is

Motor vehicle 1 designed as a passenger car. The motor vehicle 1 comprises a camera system 2. The camera system 2 in turn can be part of a

Driver assistance system 3. Furthermore, the camera system 2 comprises at least one fisheye camera 4. In the present embodiment, this includes

Camera system 2, for example, four fisheye cameras 4, which turn on

different places are distributed over the motor vehicle 1. However, for example, only two fisheye cameras 4 or three fisheye cameras 4 or more than four fisheye cameras 4 can also be provided. These are one of the

Fisheye cameras 4 in a rear area 5, one of the fisheye cameras 4 in a front area 7 of the motor vehicle 1 and the remaining two

Fisheye cameras 4 are arranged in a respective lateral area 6, in particular in an area of the side mirrors of the motor vehicle 1. In the present case, the number and arrangement of the fisheye cameras 4 of the camera system 2 are to be understood purely as examples.

In the method according to the invention for operating the fisheye camera 4 for the driver assistance system 3 of the motor vehicle 1, the fisheye camera 4 uses the fisheye camera 4 to produce a fisheye image 12 (FIG. 2) depending on at least a first image area 13 (FIG. 2) of the fisheye image 12 and a second image area 14 (Fig. 2) of the Fish eye image 12 for the driver assistance system 3 by means of the electronic

Computing device 1 1 of the fisheye camera 4 evaluated.

It is provided that the first image area 13 through a first distance area 15 (FIG. 2) to the fisheye camera 4 and the second image area 14 through a second distance area 16 different from the first distance area 15

Fisheye camera 4 are specified and the first image area 13 is compressed with a first compression algorithm 17 (FIG. 2) and the second image area 14 is compressed with a second compression algorithm 18 (FIG. 2) that is different from the first compression algorithm 17 and one as a function of the compressed first

Image area 13 and the compressed second image area 14 compressed image 19 is generated by means of the electronic computing device 11 and sent to the

Driver assistance system 3 is transmitted for evaluation.

In particular, compression 25 (FIG. 3) of the fisheye image 12 takes place as a function of the respective image area 13, 14 of the fisheye image 12, the image area 13, 14 being particularly dependent on the distance. For example, it can be provided that a close-up area of the fisheye camera 4, which is an area directly in front of the motor vehicle 1, in particular due to the arrangement on the motor vehicle 1, is compressed accordingly, since in particular little information is available here, and therefore a small loss of information during compression 25 takes place. In particular, an area further away from the fisheye camera 4 can be identified as the image area 13, 14 and this in turn can be correspondingly less compressed, in particular decompressed, that is to say a negative compression 25, so that a higher information density can be generated in this image area 13, 14. In particular, these two compressed image areas 13, 14 are then combined into the common compressed image 19 and to the

Hand over driver assistance system 3 for transmission.

The driver assistance system 3 can then, for example, process the compressed image 19 by means of the computer vision algorithm, as a result of which computing time can be saved. In particular, the computing time is saved without having to accept significant information losses. This enables a more efficient use of the electronic computing device of the driver assistance system 3.

In particular, a faster evaluation of the fisheye image 12 can thereby be carried out. Furthermore, it is possible that, for example, processors are used in the electronic computing device which are smaller Can have computing power, which for example saves space or costs with respect to the electronic computing device.

In particular, it can be provided that the first compression algorithm 17 and / or the second compression algorithm 18 are used to compress the first image area 13 and / or the second image area 14 with respect to the amount of data in the fisheye image 12. In other words, the image can be compressed in particular with respect to the respective pixels in the first image area 13 and / or in the second image area 14. A smaller amount of data is thus provided in the compressed image 19, which can then be transmitted to the driver assistance system 3.

Furthermore, it can be provided that the visually visible image can also be compressed in accordance with its size, as a result of which the amount of data which is transmitted to the driver assistance system 3 is in turn compressed accordingly.

In particular, a flowing compression 25 can take place between the first image area 13 and the second image area 14. In other words, at least at a common border area of the first image area 13 and the second

Image area 14 essentially the same compression 25 instead. In particular, linear compression 25 is thus recorded in adjacent image areas 13, 14, which is preferred.

In particular, the compressed image 19 also has the same objects 26, 27 (FIG.

5) in the area 4 of the fisheye camera 4 on how the captured

Fisheye image 12, while the compressed image 19 is significantly smaller. Furthermore, the compressed image 19 has a difference in the object size, in particular objects 26, 27 which are further away appear larger than objects 26, 27 which appear close, in particular the objects 26, 27 which are close are shown smaller , which can be particularly advantageous for object detection.

In particular, it can be provided that, in addition to generating the

compressed image 19, a fisheye correction 20 (FIG. 3) is carried out by means of the electronic computing device 11. In particular, it can be provided that the fisheye correction 20 is carried out before generating the compressed image 19. In other words, in addition to the compression 25 of the first image area 13 and the second image area 14, the image 19 is also corrected Fisheye image 12 performed. In particular, this is already done using the

electronic computing device 11 of the fisheye camera 4, so that the driver assistance system 3 can be provided with a compressed and corrected image 19. This has the particular advantage that the

Driver assistance system 3 can already evaluate the compressed and corrected image 19, as a result of which a computational capacity-saving and reliable evaluation of the fisheye image 12 can be carried out by means of the driver assistance system 3.

FIG. 2 shows a schematic view of a recording of a fisheye camera 4. In the part a) the fisheye image 12 is shown, while in the part b) the compressed image 19 is shown. In part a) it can also be seen that the first distance range 15 is specified as the close range of the fisheye camera 4 and the second distance range 16 is specified as the far range of the fisheye camera 4. In particular, FIG. 2 shows that a compression 25 of the near range is greater than a compression 25 of the far range. In particular, it can be seen that the far area is decompressed and the near area is compressed. In particular, can

For example, the far range may include a vanishing point 28 (FIG. 6) of the fisheye camera 4. In particular, no compression 25 takes place in the vanishing point 28, while the compression 25 increases with a distance from the vanishing point 28. In other words, it is provided in particular that a non-linear compression 25 is also carried out within the near range and / or the far range.

In particular, this prevents information losses from taking place in the long range, while the information density in the close range increases, for example, so that the surroundings 4 can still be reliably detected and, for example, object detection can be carried out using the driver assistance system 3.

The near area and the far area preferably have a common boundary in the fisheye image 12, so that a smooth transition from the far area to the

Close range is realized. In particular, the fisheye image 12 only has the far and near range.

Furthermore, FIG. 2 shows that at least a third image area 21 of the fisheye image 12 is compressed by means of at least a third compression algorithm 22, which is different from the first compression algorithm 17 and the second compression algorithm 18, and one as a function of the compressed first image area 13, the compressed second Image area 14 and the at least third compressed Image area 21 compressed image 19 is generated by means of the electronic computing device 10 and is transmitted to the driver assistance system 3 for evaluation.

In part b) of FIG. 2 it is shown how after using the first

Compression algorithm 17, the second compression algorithm 18 and the third compression algorithm 22, the compressed image 19 is generated. In particular, it can be seen that the far area has experienced no or a low compression 25, while the near area has experienced a strong compression 25. Furthermore, the sky area of the third image area 21 has also received a strong compression 25. The compressed image 19 is significantly reduced both in its data size and in its optical display size. In particular, this compressed image

19 then in turn transmitted to the driver assistance system 3. In particular, it can be provided, for example, that the compressed image 19 is sent to the first

Driver assistance subsystem 9 and / or is sent to the second driver assistance subsystem 10.

Fig. 3 shows a schematic block diagram of an embodiment of the

Procedure. The fisheye image 12 is made available to the electronic computing device 11. In method block 23, both compression 25 and fisheye correction 20 are carried out. For this purpose, the electronic computing device 11 can have a look-up table 24. In the look-up table 24, which can also be referred to as a look-up table, at least one extrinsic and / or intrinsic parameter of the fisheye camera 4 can be stored in the look-up table 24 of the electronic computing device 11 and the fisheye correction

20 and the compression 25 depending on the stored extrinsic and / or the stored intrinsic parameter. The parameters of, in particular, can be used as input values for the look-up table 24

Fisheye correction 20 and parameters for the compression 25 are provided. After the fisheye correction 20 and the compression 25, the compressed image 19 is then sent to the driver assistance system 3, in particular to a first driver assistance subsystem 9 and a second driver assistance subsystem 10, here, for example, using a computer vision algorithm

compressed image 19 can be evaluated.

Object detection can be carried out by means of the first vehicle subsystem 9, for example, and a parking process can be supported by means of the second driver assistance subsystem 10. In particular, it can be provided that, depending on a function of the

Driver assistance system 3 or the respective driver assistance subsystem 9, 10 a parameter of the first compression algorithm 17 for compressing 25 the first image area 13 and / or a parameter of the second compression algorithm 18 for compressing 25 the second image area 14 by means of the electronic

Computing device 1 1 is selected.

Fig. 4 shows a schematic view of an embodiment of the method. 4 shows that the fisheye image 12 can be compressed into different compressed images 19 ″, 19 ″, 19 ″ by means of the compression 25. In particular, FIG. 4 shows that the fisheye image 12 can be compressed into a first compressed image 19 ″ and / or into a second compressed image 19 ″ and / or into a third compressed image 19 ″ ”?

Furthermore, FIG. 4 shows that the first image area 13 and / or the second one, in particular by means of a non-linear hyperbolic projection algorithm 33 (FIG. 7)

Image area 14 can be compressed. In other words, the first compression algorithm 17 is a non-linear hyperbolic

Projection algorithm. In particular, it can also be the second

Compression algorithm 18 around a non-linear hyperbolic

Act projection algorithm 33. In other words, it is provided that the compression 25 is therefore not performed linearly in the first image area 13 and / or in the second image area 14. The respective strength of the compression 25, in particular pixel by pixel, can thus be realized and reproduced by means of the non-linear hyperbolic projection algorithm 33. By means of the non-linear hyperbolic

Projection algorithm 33, a reliable compression algorithm 17, 18, 22 can be provided, by means of which computing effort can be reduced and the compressed image 19 ″, 19 “, 19 ″” can be evaluated by means of the driver assistance system 3.

4 furthermore shows that a parameter of the first compression algorithm 17 for compressing 25 the first image area 13 and / or a parameter of the second compression algorithm 18 for compressing 25 the second image area 14 by specifying an image size G and / or a processing time of the compressed image 19 ', 19 ", 19'" for driver assistance system 3 can be specified by driver assistance system 3. In other words it can be provided, for example, that the Driver assistance system 3 only requires an image size G of a predetermined number of pixels for processing, in which case the compression 25 of the first image area 13 and of the second image area 14 can then be carried out as a function of this image size G. Furthermore, it can be provided, for example, that the driver assistance system 3 only evaluates a predetermined processing time for evaluating the transmitted image 19 ', 19 ", 19'" and, depending on this, the corresponding data size is again calculated and the fisheye image 12 is compressed accordingly by this processing time to be able to achieve. This makes it possible for the driver assistance system 3 to do this reliably and as a function of a desired processing time and / or a predetermined image size G

compressed image 19 ″, 19 ″, 19 ″ ”can be adjusted.

FIG. 4 further shows that the first compression algorithm 17 and the second compression algorithm 18 perform a compression 25 of the image 12 between 20% (first compressed image 19 ″) and 70% (third compressed image 19 ″ ”). For example, it is further shown in FIG. 4 that the second compressed image 19 “has undergone a compression 25 of 50%.

FIG. 5 shows a further schematic view of a fish eye image 12 in FIG

Fisheye camera 4. In part a), the fisheye image 12 can be seen, which was recorded by the fisheye camera 4. In part b) the second compressed image 19 "is shown with a compression 25 of 50%. In part c) the third compressed image 19 "is shown with a compression 25 of 70%.

In particular, it is provided that the first image area 13 and / or the second image area 14 is determined as a function of a detection frequency of an occurrence of the specific object 26, 27. For example, it can be provided that a pedestrian is specified as the specific object 26, 27. In particular, pedestrians who are of importance for the driver assistance system 3 are in the vicinity 8 of the fisheye camera 4. In particular, the distance range 15, 16 for the first image area 13 and / or for the second image area 14 in

Dependency of this object 26, 27 can be made. If, for example, the object 26, 27 is specified as a motor vehicle, the respective distance range 15, 16 can be enlarged, since in particular the detection of motor vehicles in the

The immediate vicinity of the fisheye camera 4 is of less importance than the detection of motor vehicles in a distant environment 8

Fisheye camera 4. In particular, the definition of the first image area 13 and in particular the associated distance area 15 and / or the second image area 14 and in particular the associated second distance area 16 are defined by the respective objects 26, 27 to be detected or by the function of the driver assistance system 3 or the driver assistance subsystem 9, 10. In particular, it is provided as a frequency of detection, for example, that the occurrence of a pedestrian in the near range is greater than in the far range, while, for example, the occurrence of the motor vehicle in the far range is higher than in

Close range. This enables a situation-dependent compression 25 of the fisheye image.

Fig. 6 shows a schematic view of another embodiment of the

Procedure. Vanishing point 28 is shown in particular in fisheye image 12.

In particular, FIG. 6 shows that compression 25 can be carried out both in the horizontal direction and in the vertical direction. In particular, a vertical compression 25 ″ and a horizontal compression 25 ″ can be carried out, which is shown by the arrows in the fisheye image 12 and in the compressed image 19.

In particular, the compression 25 follows a non-linear hyperbolic projection function.

7 shows a schematic diagram of an embodiment of the method.

In particular, it is shown that no compression 25 takes place at vanishing point 28. The further away from vanishing point 28, the greater the compression 25.

For example, the graph 29 shows a compression 25 with 24%, the graph 30 a compression 25 with 32.5%, the graph 31 a compression 25 with 48.8% and the graph 32 a compression 25 with 72%. In particular, the compressed image 19 is plotted on the ordinate O and the fisheye image 12 is plotted on the abscissa A. In particular, these are corresponding distances.

Overall, the invention shows a distance-dependent image region compression for efficient evaluation of an image 12.

The compression between 20% and 70% as described above is only an example. Compression up to 90% is possible.

Claims

1. Method for operating a fisheye camera (4) for a

Driver assistance system (3) of a motor vehicle (1), in which the

Fisheye camera (4) a fisheye image (12) is recorded and the fisheye image (12) in dependence on at least a first image area (13) of the fisheye image (12) and a second image area (14) of the fisheye image (12) for the

Driver assistance system (3) is evaluated by means of an electronic computing device (11) of the fisheye camera (4),

characterized in that

the first image area (13) through a first distance area (15)

Fisheye camera (4) and the second image area (14) through a second distance area (16) different from the first distance area (15)

Fisheye camera (4) and the first image area (13) with a first compression algorithm (17) and the second image area (14) with a different from the first compression algorithm (17)

Compression algorithm (18) and a compressed image (19 ″, 19 ″, 19 ″ “) depending on the compressed first image area (13) and the compressed second image area (14) by means of the electronic

Computing device (1 1) is generated and transmitted to the driver assistance system (3) for evaluation.

2. The method according to claim 1,

characterized in that

in addition to generating the compressed image (19 ″, 19 “, 19 ″”)

Fisheye correction (20) is carried out by means of the electronic computing device (1 1).

3. The method according to claim 2,

characterized in that

before the creation of the compressed image (19 ″, 19 “, 19 ″”)

Fisheye correction (20) is performed.

4. The method according to any one of the preceding claims, characterized in that

the first distance range (15) is specified as the close range of the fisheye camera (4) and the second distance range (16) as the far range of the

Fisheye camera (4) is specified.

5. The method according to claim 4,

characterized in that

a compression (25 ″, 25 “) of the near range is greater than a compression (25’, 25 “) of the far range.

6. The method according to any one of the preceding claims,

characterized in that

the first image area (13) and / or the second image area (14) are compressed by means of a nonlinear hyperbolic projection algorithm (33).

7. The method according to any one of the preceding claims,

characterized in that

at least a third image area (21) of the fisheye image (12) is compressed by means of at least one third compression algorithm (22), which is different from the first compression algorithm (17) and the second compression algorithm (18), and one depending on the compressed first image area (13 ), the compressed second image area (14) and the at least

compressed third image area (21) compressed image (19 ″, 19 “, 19 ″”) is generated by means of the electronic computing device (1 1) and sent to the

Driver assistance system (3) is transmitted for evaluation.

8. The method according to any one of the preceding claims,

characterized in that

depending on a function of the driver assistance system (3), a parameter of the first compression algorithm (17) for compressing (25 ″, 25 “) the first

Image area (13) and / or a parameter of the second compression algorithm (18) for compressing (25 ', 25 ") of the second image area (14) is selected by means of the electronic computing device (1 1).

9. The method according to any one of the preceding claims, characterized in that

at least one extrinsic and / or one intrinsic parameter of the

Fisheye camera (4) in a lookup table (24) of the electronic

Computing device (1 1) can be stored and a fisheye correction (20) is carried out depending on the stored extrinsic and / or the stored intrinsic parameter.

10. The method according to any one of the preceding claims,

characterized in that

the first image area (13) and / or the second image area (14) depending on one

Detection frequency of an occurrence of a specific object (26, 27) can be determined.

1 1. The method according to any one of the preceding claims,

characterized in that

a parameter of the first compression algorithm (17) for compressing (25 ', 25 ") the first image area (13) and / or a parameter of the second

Compression algorithm (18) for compressing (25 ″, 25 “) the second

Image area (14) by specifying an image size (G) and / or

Processing time of the compressed image (19 ″, 19 “, 19 ″“) for the

Driver assistance system (3) can be specified by the driver assistance system (3).

12. The method according to any one of the preceding claims,

characterized in that

by means of the first compression algorithm (17) and the second

Compression algorithm (18) a compression (25 ', 25 ") of the fisheye image (12) between 20% and 70% is carried out.

13. Electronic computing device (1 1) which is designed to carry out the method according to one of claims 1 to 12.

14. fisheye camera (4) with an electronic computing device (1 1)

Claim 13.

15. Driver assistance system (3) with a fisheye camera (4) according to claim 14.