HK1237111A1 - Determination of at least one feature of a vehicle - Google Patents

Description

Determining at least one characteristic of a vehicle

Technical Field

The invention relates to an arrangement for determining at least one feature of a vehicle moving along a traffic lane, having a light-sensitive sensor for recording an image of the vehicle and having an evaluation device for determining the at least one feature on the image, and a method for determining the at least one feature of a vehicle moving along a traffic lane by means of such an arrangement, wherein the image of the vehicle is recorded by means of a line-scanning camera, and wherein the image is analyzed to determine the at least one feature of the vehicle.

Background

Arrangements and methods for recording vehicles moving along a traffic lane by means of at least one camera and/or other devices for detecting the contour or spatial structure of the vehicle and, if necessary, the speed of the vehicle are known, for example, from different applications for monitoring or traffic detection.

For example, an automatic control system for road tolling, which is based on communication between the vehicle and a corresponding communication device at a toll station, requires the presence of a corresponding communication device in each vehicle paying for road tolling, from which communication device basic data for the vehicle identified by the vehicle can be called, and where used road sections should also be recorded and/or transmitted. The system must be able to clearly identify each vehicle that has to pay a charge as such a vehicle, i.e. such a vehicle that does not have to pay a charge should not be detected, and on the other hand all vehicles that have to pay a charge are detected as such a vehicle. A method and a device for recording moving vehicles are provided, which can automatically distinguish between charged vehicles and uncharged vehicles in order to thereby prevent unnecessary recording of large amounts of data, according to what is proposed in published document DE 10148289a1, the recording and tracking of vehicles is performed by means of a LIDAR-system (LIDAR-system) before recording contours and/or spatial structures, wherein the path and speed of the vehicle are estimated from the LIDAR-data (LIDAR-data) and subsequently assigned to the determined contour and structure data. The device has a lidar system that is aligned so that it records a vehicle at a distance in front of the recording location of the camera and/or other device. Furthermore, a central processing unit is provided which is connected to the LIDAR (LIDAR) and predicts the path and speed of the vehicle from the LIDAR data, corrects these data if necessary and performs an assignment of data of other devices and/or recorded video images based on these data. A disadvantage is that in the combination of a lidar system and a video imaging device the respective results have to be safely and reproducibly assigned to each other, which is cumbersome. Furthermore, in the evaluation of video images, the problem is presented that the evaluation becomes difficult due to perspective distortion and distortion itself and on the other hand due to the reduced light yield with increasing distance from the camera, which results in significantly weaker illumination of the area of the vehicle arranged further away from the camera. This is very important for longer vehicles, such as trucks.

Disclosure of Invention

The object of the present invention is to provide a simplified arrangement and method for determining at least one characteristic of a vehicle moving along a traffic lane, which enables reliable or improved characteristic recognition compared to the prior art, for example for automatically distinguishing between vehicles requiring payment of a fee and vehicles not requiring payment of a fee.

This object is achieved by an arrangement and a method according to the independent claims. Preferred embodiments and advantageous developments are given in the respective dependent claims.

The arrangement according to the invention for determining at least one characteristic of a vehicle moving along a traffic lane has a light-sensitive sensor for recording an image of the vehicle and has an evaluation device for determining the at least one characteristic on the image. According to the invention, the light-sensitive sensor is implemented as a line-scanning camera, wherein the line-scanning camera is preferably arranged laterally next to the traffic lane. Furthermore, it is provided that exactly one line scan camera is used as the photosensitive sensor. Alternatively, several (in particular two) line scan cameras may be provided as photosensitive sensors.

Furthermore, according to the invention, the number of lines per unit time of the image is adapted to the speed of the vehicle. The individual recordings of the line scan camera are designated as lines. According to a first preferred variant, the number of lines per unit time of the image is adapted to the speed of the vehicle, so that the sampling rate of the line scan camera is adapted to the speed of the vehicle. The sampling rate specifies the frequency at which the line scan camera records a line, i.e., the line recorded per unit time. A second preferred variant of adapting the number of lines per unit time of the image to the speed of the vehicle is that the post-correction of the image is performed in dependence on the speed of the vehicle. The number of lines per unit time can likewise be varied by stretching or compressing the image formed by the lines arranged one after the other. Thus, advantageously, image distortions due to the speed of the vehicle can be corrected such that the image truly shows that the aspect ratio of the image is faithful to the prototype. The velocity can be measured separately or can be determined by means of a distorted image as will be explained later. Basically, a combination of the two variants is also possible, namely an adapted image sampling rate of the camera and a post-correction of the image, for example to minimize the error margin.

Of course, it is generally possible to record vehicles moving in a roadway in an image with a common surface camera, for example with an electromagnetic coil as a trigger for the camera, however, such an image is distorted in perspective in both image dimensions. The use of a line scan camera has the advantage that the distance from the recorded part of the vehicle to the line scan camera remains substantially constant, so that neither perspective distortions in the direction of the additional recording one after the other nor different light yields due to the distance between vehicle and camera occur. For this reason, it is particularly advantageous to use a line scan camera for the arrangement laterally beside the traffic lane. The lateral positioning next to the carriageway makes it possible to particularly effectively recognize many features of the vehicle in the image, in particular the length of the vehicle, the number of axles, the feature shape and the lateral text which can contribute to the identification. Many of these features are not visible, for example, at locations above the roadway. In this case, the images are constructed from each other after recording by a line-scan camera operating at a predetermined, if necessary adaptive, sampling rate. Another advantage is that the line scan camera can be used with an evaluation device for image processing without the need for additional measuring devices. The beginning and the end of the vehicle can be detected by means of analyzing the continuously recorded line scans, so that advantageously no triggering of the recording to generate a complete image of the vehicle is necessary. Since the sampling rate of the line scan camera is known, the speed of the vehicle can also be determined by means of a reference on the image, which is displayed in dependence on the speed of the vehicle, compressed or stretched in the horizontal axis of the image corresponding to the time axis with respect to the true original aspect ratio of the vehicle, for a constant sampling rate. Thus, for example, a known length of an object on a vehicle may be used as a reference. Particularly advantageously, each wheel can be used as a reference. Such as a wheel of a round vehicle with sufficient accuracy, the speed of the vehicle may be determined by calculating the ellipticity, depending on the speed of the vehicle, shown as an ellipse due to compression or stretching in height or width. Ellipticity, also known as eccentricity, states the distance of the focus of an ellipse from its center point.

Laterally beside the traffic lane, it is to be understood in the sense of the present invention that the line scanning camera is not arranged vertically above the traffic lane, but extends substantially horizontally. However, it is obvious that the line scanning camera may also be arranged beside the traffic lane at a predetermined height above the level of the traffic lane.

A line scan camera in the sense of the present invention is a camera type having only one light sensitive line compared to a two-dimensional sensor having a plurality of lines, so-called line sensor. The line sensor is a light or radiation sensitive detector, typically a semiconductor detector comprising a one-dimensional field photodetector or other detector element.

According to an embodiment, the line scan camera is aligned obliquely to the traffic lane. In the sense of the present invention, the alignment of a line scan camera is the direction in which the light sensitive lines are aligned to receive light. The vehicle moves in the direction of travel along the lane. A line scan camera arranged beside the traffic lane can thus be aligned at right angles to the vehicle and the traffic lane and record the side thereof. If the angle between the traffic lane and the line scanning camera is not equal to 90 °, the line scanning camera is aligned obliquely to the traffic lane in the sense of the invention. In the preferred case of a line-scan camera, which is aligned obliquely at an acute angle to the traffic lane, opposite to the direction of travel of the vehicle, firstly the front of the vehicle and secondly the side of the vehicle are recorded, which has the advantage that the vehicle registration and possibly the hazardous-article label can be detected on the front of the vehicle. In the case of an oblique alignment of the line scanning camera at an obtuse angle to the traffic lane in the direction of travel of the vehicle, the line scanning camera is also aligned obliquely to the traffic lane in the sense of the invention. In this preferred case, firstly the side of the vehicle and secondly its rear side are recorded, which likewise enables a vehicle registration marking by means of a rear-side number plate. It is particularly preferred to provide that the line scanning camera is aimed at an angle of approximately 30 ° to the traffic lane opposite to the direction of travel of the vehicle. Thus, the angle of the line scan camera to a straight line in front of the vehicle is approximately 60 °.

The alignment of the line scanning camera does not mean, in the sense of the present invention, the alignment of the line sensor itself, which is usually arranged vertically, i.e. substantially at right angles to the plane formed by the traffic lane. It is also generally possible for the line scan sensor to be positioned at an angle significantly below 90 deg. to the vertical.

According to a further embodiment of the arrangement, an illumination device is provided, wherein the illumination device is aligned at an angular distance of at least 7 ° from the alignment of the line scan camera. The lighting device is used for illumination in low light conditions, especially at night. An angular distance of at least 7 ° from the alignment of the line scanning camera means that light emitted from the lighting device is reflected by the vehicle and light detected by the line scanning camera is deflected in reflection by at least 7 °. Therefore, excessive incident light that obstructs evaluation of the number plate having the retroreflective surface can be advantageously prevented. For illumination, visible light as well as infrared radiation is suitable.

According to another preferred embodiment of the arrangement, a control of the line scanning camera is provided, wherein the sampling rate of the recorded lines is adjustable by the control. Furthermore, preferably, a speed measuring device is provided for measuring the speed of the vehicle when recording the image. Thus, a particularly advantageous embodiment of the arrangement can be achieved, wherein the sampling rate is adapted by controlling the vehicle speed measured by the speed measuring device. Thus, compression or stretching of the image on the horizontal axis of the image is at least prevented as far as possible, so that post-correction, for example by means of the ovality of the shown wheel, can be omitted or must at least be performed less extensively. In an alternative embodiment, the measurement of the speed may be verified by a speed measurement device with a constant sampling rate of the line scan camera by means of a speed that can be determined mathematically from the image. The speed of the vehicle, which is determined mathematically and/or measured from the image, may additionally or alternatively be used to correct the image in a later stage where the image is compressed or stretched, so that distortions due to the speed of the vehicle are balanced.

According to the method of the invention for determining at least one characteristic of a vehicle moving along a traffic lane, in particular provided by means of the arrangement described above, an image of the vehicle is recorded with a line-scanning camera and the image is evaluated to determine the at least one characteristic of the vehicle. One advantage of this method is that the identification of the vehicle can be achieved only by means of the line scan camera and the evaluation of the images. To this end, in particular individually recorded lines or line scans are appended one after the other on the horizontal axis of the image, so that the horizontal axis of the image corresponds to the time axis. The linear markers generated along the horizontal axis of the image thus show a time-constant image from which it can be deduced that no vehicle passes through the image point of the line-scanning camera. Such a horizontal linear marking, or in particular the end of a horizontal linear marking, can advantageously be used for automatic detection of a moving vehicle.

Furthermore, according to the invention, the number of lines per unit time of the image is adapted to the speed of the vehicle. According to a first preferred variant, the number of lines per unit time of the image is adapted to the speed of the vehicle, so that the sampling rate of the line scan camera is adapted to the speed of the vehicle. A second preferred variant is that the image is corrected at a later stage according to the speed of the vehicle in order to adapt the number of lines per unit time of the image to the speed of the vehicle. Particularly preferably, the speed of the vehicle is measured during the recording of the image, and the image is corrected as a function of the speed by means of stretching or compressing on the horizontal axis of the image. Furthermore, it is preferred that the speed of the vehicle is determined mathematically by means of a distortion of the image, and that the image is corrected by means of the mathematically determined speed by stretching or compressing on the horizontal axis of the image.

According to a preferred embodiment, during the evaluation, an envelope of the image enclosing the vehicle is generated. Thus, at least one of the height, length, and width of the dimension of the vehicle may be calculated taking into account the sampling rate of the line scan camera and the speed of the vehicle. In this case, the speed of the vehicle may be determined by stretching or compressing the image on its horizontal axis relative to the true native aspect ratio on the vehicle, or may be determined by a separate speed measurement.

According to a further preferred embodiment, it is provided that the number and/or position of the wheels and/or axles of the vehicle can be determined by means of finding a circular object during the image evaluation. Furthermore, the ellipticity is determined from circular objects detected during image analysis. Particularly preferably, the correction of the image is performed by means of stretching or compressing on the horizontal axis of the image taking into account the determined ellipticity. According to a further preferred embodiment, it is provided that the speed of the vehicle is calculated taking into account the ellipticity and the sampling rate of the line-scan camera. This calculated speed is another characteristic of the vehicle, which can advantageously be determined by means of the line scan camera as the only measuring device. Alternatively, the calculated speed may be used to check the speed measured by a separate speed measuring device.

According to a further preferred embodiment, it is provided that the classification and/or identification of the vehicle is performed by means of at least one of the following features received by means of the analysis of the image: the size of the vehicle, the number of axles, vehicle registration, hazardous article tags, shape of the front windshield and/or side windows, and feature shapes.

Drawings

In the following, the invention is described in more detail by means of embodiments with reference to the accompanying drawings. These explanations also relate to the arrangement and the method according to the invention. These explanations are merely exemplary and do not limit the general concept of the invention.

Shows that:

figure 1 a first schematic view of an arrangement according to the invention;

FIG. 2 is a second schematic view of the arrangement of FIG. 1;

FIG. 3 is an image of a vehicle produced according to the method of the present invention.

Detailed Description

Fig. 1 and 2 show two different views of an arrangement according to the invention for determining at least one characteristic of a vehicle 2 moving along a traffic lane 1, which views are described together below. Fig. 1 shows a top view of the traffic lane 1 and the vehicle 2, while fig. 2 shows a side view of the vehicle 2. The arrangement according to the invention comprises a light-sensitive sensor 3 for recording an image of the vehicle 2 and an evaluation device 4 for determining at least one feature on the image, wherein the evaluation device 4 can also be arranged at a distance from the sensor 3. According to the invention, the photosensitive sensor 3 is implemented as a line scan camera 3, wherein the line scan camera 3 is arranged laterally beside the carriageway 1.

The arrangement according to the invention is particularly advantageous for a method for axle classification or is suitable for calculating the number of axles of a vehicle 2 in flowing traffic. This method is generally used for the classification of vehicles 2, in particular for determining salient features while moving, such as vehicle manufacturers or dangerous goods labels. Additionally, an image of the vehicle registration suitable for the automatic registration reading process is recorded with the same sensor 3. Furthermore, an image of the vehicle 2 is generated as an overview of the entire scene. With the arrangement according to the invention and the method according to the invention, four objects in the field of traffic monitoring can be carried out by means of only one line scanning camera 3 alongside the traffic lane 1. The line scanning camera 3 is arranged laterally next to the traffic lane 1, so that the front 5 of the vehicle and the side 6 of the vehicle 2 driving through are cyclically recorded. In case the natural ambient light is not sufficient, especially at night, the lighting device 7 is preferably directed onto the vehicle 2. In this case, the angle between the line scanning camera 3 and the vehicle 2 is generally freely selectable and can be adapted according to the application. In general, it is advantageous to have an angle as close to 90 ° as possible with the area to be monitored. As in the illustrated embodiment, the front 5 of the vehicle and the side 6 of the vehicle should be recorded, an angle of approximately 60 ° to the front 5 is recommendable, since this contains more detailed information than the side 6, from which the light is reflected into the line scanning camera 3 at a correspondingly sharper angle of 30 °. As can be seen from fig. 2, the light starting from the lighting device 7 and schematically represented by the arrow marked L and reflected by the vehicle 2 preferably falls into the line scanning camera 2 at an angle α of more than 7 °. At angular distances between the lighting device 7 and the camera 3 of more than 7 °, overexposure due to the so-called retro-reflection effect of the number plate is advantageously prevented.

With the method according to the invention an image of the vehicle 2 is generated and thus a simple image processing of this same image is enabled, since no partial images have to be put together or covered. Advantageously, no distortion on the side 6 of the vehicle 2 is produced on the horizontal axis of the image, since the distance between the camera 3 and the vehicle 2 remains substantially constant, except for deviations due to turning movements of the vehicle 2 or unevenness of the side 6 of the vehicle 2. Further, no falling line is generated in the image on the side face 6 of the vehicle 2. The method according to the invention is also suitable for determining the speed of the vehicle 2. From which the speed of the vehicle 2 can be derived, depending on how elliptically the wheels 8 are represented on the image (only schematically shown here).

All the features described here can be identified by an arrangement according to the invention with only one sensor 3. In fig. 1 and 2, the sensor 3 may be placed slightly above the road surface. It is also possible to mount the sensor 3 on a bridge (not shown), however, in any case the mounting of the sensor 3 is laterally offset in the region to be monitored, i.e. the traffic lane 1.

A laser-based speed sensor 9, such as a LIDAR-sensor (LIDAR-sensor) used in speed measurement devices, may preferably be provided as an additional detector of the vehicle 2. The speed sensor 9 and the line scanning camera 3 are arranged at the same position laterally beside the road 1. At the point in time of detection, the speed sensor 9 preferably delivers the following data of the vehicle: the position, distance and speed of the vehicle 2 relative to the sensors.

Fig. 3 shows an image of the vehicle 2 produced by the method according to the invention. In contrast to an ordinary camera (so-called still camera or area scanning camera) which shows that both coordinate axes are distorted by perspective in the image spatial position, when the line scanning camera 3 having the arrangement according to the present invention is used, the horizontal axes of the images are replaced by a time series of line recordings produced one after the other along the time axis designated with t, and therefore the horizontal axes of the images are free from perspective distortion. The distance from one pixel to the next in the direction of the t axis corresponds to the sampling rate or line frequency of the line scan camera 3 and is preferably adapted to the speed of the vehicle 2 determined by the speed sensor 9 such that no squeezed or stretched image is produced in the direction of the t axis. The vertical axis of the image in the direction of the axis designated by Y corresponds to the alignment of the line scanning sensor 3.

The line scanning camera 3 is arranged at a fixed position and is not moved. A stationary object is thus only represented by horizontal lines of low dynamics, which are likely to be due to pixel noise or shading. Fig. 3 shows an image of the moving truck 2. The contour or envelope H of the vehicle 2 can advantageously be determined by means of a simple edge filter. Even shadows throughout the image can be easily identified and do not lead to erroneous assessments, as is often the case when recording with a still camera, since a time reference cannot be generated. Geometric information, such as length and/or height, can advantageously be determined by means of the edge image. If the position of the axle of the vehicle is known, it can be advantageously excluded that the front side of the vehicle 2 is additionally measured. The measurement of the length of the vehicle is started, for example, from the first determined axle. Depending on the identified vehicle type (e.g. truck or bus), the length may be corrected by an averaging factor.

Other sensors used to monitor traffic, such as lidar scanners, typically operate at a sampling rate of 100 to 200Hz and therefore do not deliver detailed data (e.g., the wheels are displayed with only about three pixels), which do not allow for extensive evaluation compared to the images produced by line-scan cameras. Line scan cameras (as they are used according to the invention) can operate at line frequencies up to 80kHz, thus advantageously providing high level possibilities for using data. The invention enables all the data necessary for road toll collection to be generated with the line scan camera 3 as the only sensor and with only one data recording cycle. These are in particular: registration of vehicle number plate images, registration of overall images, classification of the vehicle based on its axle, and classification of the vehicle geometry. Furthermore, it is possible to determine the hazardous item label and the vehicle manufacturer. Since all data can be generated in one step, a consistent integer data set is generated. Erroneous distribution of data from several sensors can be advantageously prevented.

List of reference marks

1-way lane

2 vehicle

3 line scanning camera

4 evaluation device

5 front of vehicle

6 side of vehicle

7 Lighting device

8 wheel

9 speed measuring device

L arrow head

Angle alpha

t time axis, horizontal axis of image

Y image vertical axis

H envelope

Claims (23)

1. An arrangement for determining at least one feature for identifying a vehicle moving along a traffic lane, having at least one light-sensitive sensor for recording an image of the vehicle and having an evaluation device for determining at least one feature on the image,

wherein the photosensitive sensor is implemented as a line scan camera and wherein the number of lines per unit time of the image is adapted to the speed of the vehicle.

2. An arrangement as set forth in claim 1, wherein,

it is characterized in that the preparation method is characterized in that,

the number of lines per unit time of the image is adapted to the speed of the vehicle such that the sampling rate of the line scan camera is adapted to the speed of the vehicle.

3. The arrangement of any one of claims 1 or 2,

it is characterized in that the preparation method is characterized in that,

the number of lines per unit time of the image is adapted to the speed of the vehicle such that a post-correction of the image is performed in dependence on the speed of the vehicle.

4. The arrangement of any one of the preceding claims,

it is characterized in that the preparation method is characterized in that,

the line scan camera is arranged laterally beside the traffic lane.

5. The arrangement of any one of the preceding claims,

it is characterized in that the preparation method is characterized in that,

the line scanning camera is aligned obliquely to the traffic lane, wherein the line scanning camera is specifically aligned at an angle of about 30 ° to the traffic lane opposite to the direction of travel of the vehicle.

6. The arrangement of any one of the preceding claims,

it is characterized in that the preparation method is characterized in that,

an illumination device is provided, wherein the illumination device is aligned at an angular distance of at least 7 ° from the alignment of the line scanning camera.

7. The arrangement of any one of the preceding claims,

it is characterized in that the preparation method is characterized in that,

a speed measurement device is provided to measure the speed of the vehicle when recording the image.

8. The arrangement of any one of claims 2 to 7,

it is characterized in that the preparation method is characterized in that,

a control for the line scan camera is provided, wherein the sampling rate of the recorded lines is adjustable by the control.

9. An arrangement as set forth in claim 8,

it is characterized in that the preparation method is characterized in that,

the sampling rate is adapted to the measured speed of the vehicle by means of the control.

10. The arrangement of any one of the preceding claims,

it is characterized in that the preparation method is characterized in that,

two line scan cameras are provided as photosensitive sensors.

11. Method for determining at least one characteristic for identifying a vehicle travelling along a traffic lane, wherein an image of the vehicle is recorded with at least one line scanning camera and the image is analyzed to determine at least one characteristic of the vehicle, wherein the number of lines per unit time of the image is adapted to the speed of the vehicle.

12. The method of claim 11, wherein the first and second light sources are selected from the group consisting of,

it is characterized in that the preparation method is characterized in that,

the number of lines per unit time of the image is adapted to the speed of the vehicle such that the sampling rate of the line scan camera is adapted to the speed of the vehicle.

13. The method according to any one of claims 11 or 12,

it is characterized in that the preparation method is characterized in that,

the number of lines per unit time of the image is adapted to the speed of the vehicle such that the image is post-corrected according to the speed of the vehicle.

14. The method of any one of claims 11 to 13,

it is characterized in that the preparation method is characterized in that,

the recorded lines are appended one after the other along the horizontal axis of the image, where linear markers appearing along the horizontal axis of the image are used to detect the vehicle.

15. The method of claim 14, wherein the first and second light sources are selected from the group consisting of,

it is characterized in that the preparation method is characterized in that,

an envelope of an image enclosing the vehicle is generated.

16. The method of claim 15, wherein the first and second light sources are selected from the group consisting of,

it is characterized in that the preparation method is characterized in that,

at least one of a height, a length, and a width of a dimension of the vehicle is calculated from the envelope in consideration of a sampling rate of the line scan camera and a speed of the vehicle.

17. The method of any one of claims 11 to 16,

it is characterized in that the preparation method is characterized in that,

the number and/or position of the wheels and/or axles of the vehicle is determined by means of detecting circular objects during the analysis of the image.

18. The method of any one of claims 11 to 17,

it is characterized in that the preparation method is characterized in that,

the ellipticity of a circular object detected during the analysis of the image is determined.

19. The method of claim 18, wherein the first and second portions are selected from the group consisting of,

it is characterized in that the preparation method is characterized in that,

taking into account the ellipticity, the image is corrected by means of stretching or compressing along the horizontal axis of the image.

20. The method of claim 19, wherein the first and second portions are selected from the group consisting of,

it is characterized in that the preparation method is characterized in that,

calculating a speed of the vehicle taking into account the ellipticity and a sampling rate of the line scan camera.

21. The method of any one of claims 11 to 20,

it is characterized in that the preparation method is characterized in that,

the speed of the vehicle is measured when the image is recorded, and the image is corrected by means of stretching or compressing along a horizontal axis of the image according to the measured speed.

22. The method of any one of claims 11 to 21,

it is characterized in that the preparation method is characterized in that,

the classification of the vehicle is performed by means of at least one of the following features received by analyzing the image: the size of the vehicle, the number of axles, vehicle registration, hazardous article tags, shape of the front and/or side windows, and feature shape.

23. The method of any one of claims 11 to 22,

it is characterized in that the preparation method is characterized in that,

an arrangement according to any of claims 1 to 10 is used.