JP2850890B2 - Composite sensor type vehicle detector - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a vehicle sensor,
In particular, the present invention relates to a vehicle detector that detects a vehicle and provides traffic information.

[0002]

2. Description of the Related Art Japanese Patent Application Laid-Open No. 5-296 is an example of the prior art.
There is an invention of an extraneous object detection device described in Japanese Patent Publication No. 766 (hereinafter, referred to as Publication 1). The invention described in Publication 1 detects an extraneous object by an imaging unit that images visible light and infrared light.

FIG. 6 is a block diagram showing a configuration of an obstacle detection device described in Japanese Patent Application Laid-Open Publication No. HEI 7-163131. The obstacle detection device shown in FIG. 6 includes a CCD having (m × n) pixels,
The configuration includes infrared cut filter means and area switching means for selecting a signal in an area having the infrared cut filter means and a signal in other areas.

There is also an invention of an image-type vehicle sensing method described in Japanese Patent Application Laid-Open No. 5-298594 (hereinafter referred to as Japanese Patent Publication No. 2). The invention described in Publication 2
The road image data (luminance data group Dt) is obtained, and the reference road surface level data (Dr) corresponding to the case where there is no vehicle and only the road surface is held. The luminance data group Dt is lower than Dr and is constant. A feature that hardly changes at the level is detected to distinguish a vehicle from a shadow. FIG. 7 is a flowchart showing a shadow determination process of the image-based vehicle sensing method described in Japanese Unexamined Patent Application Publication No. 2000-209873.

[0005]

A first problem is that it is difficult to reliably determine that no vehicle exists when acquiring road surface data when no vehicle exists. .

[0006] The reason is that, in the prior art described in Japanese Patent Laid-Open Publication No. 2000-209, when distinguishing a vehicle from a shadow, in order to remove the influence of the surrounding environment, there is a case where the road image data and the vehicle do not exist. This is because the difference from the road surface data is used.

The second problem is that even if a specific portion of the CCD composed of (m × n) pixels is formed with a portion with and without an infrared cut filter, the portion between the visible light region and the thermal infrared region is not affected. That is, you cannot get light.

The reason is that, in the conventional technique described in Japanese Unexamined Patent Publication (Kokai) No. 1 (1994), the presence or absence of an infrared cut filter is used to convert light in two different wavelength bands into an electric signal.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned conventional problems and to eliminate the influence of objects other than vehicles without being affected by shadows of vehicles and buildings, road surface conditions, and surrounding environments such as objects other than vehicles. An object of the present invention is to provide a composite sensor type vehicle sensor that distinguishes a running or stopped vehicle and improves vehicle detection accuracy.

[0010]

In order to achieve the above object, a composite sensor type vehicle sensor according to the present invention is a vehicle sensor for detecting a vehicle by imaging a road surface and has a sensitivity in a visible light region. And a visible region imaging device (1) for capturing an image in the visible region.
1) an infrared imaging device (12) having sensitivity in a thermal infrared region and capturing an image in the infrared region; an illuminometer (21) and a thermometer (22) for measuring an environment near a road surface. Environment measuring device (20), the visible region image data output from the visible region imaging device (11), and the infrared region imaging device (1).
And 2) processing the image data and the environmental data the environmental measuring devices (20) to the output of the infrared region output from the signal processing device for detecting a vehicle (30), the environment measuring device road
A first method of acquiring the environmental data in the vicinity at predetermined time intervals;
Means, and the visible region imaging apparatus takes an image of the visible region.
A second means for imaging; and
Third means for capturing an image of a line area, and the signal processing device
Is a road surface using the environmental data and the image data in the visible region.
Means for detecting the surrounding environment of the vehicle, and the fourth means
Weather condition is determined based on the detected surrounding environment
Fifth means and normal weather conditions in the fifth means
If it is determined that there is, use the image data in the visible range.
Sixth means for detecting a first candidate for a vehicle,
Means for detecting the first candidate of the vehicle
If it is determined that the first candidate of the vehicle has not been detected,
Seventh means returning to the first means, and the seventh means
The first candidate of the vehicle is detected or the fifth hand is detected.
If it is determined in the step that the weather is bad,
Detects second vehicle candidates using infrared image data
Eighth means, and the second means of the vehicle in the eighth means.
It is determined whether or not a candidate has been detected, and a second candidate for the vehicle is determined.
If no is detected, the ninth step returns to the first means.
Detecting a second candidate of the vehicle in the step and the ninth means
In the event that the vehicle is
Output means.

[0011]

In the above-mentioned composite sensor type vehicle sensor according to the present invention, the fourth means may be configured such that the average value and the standard deviation of the road surface luminance data are obtained by using the visible region image data acquired by the visible region imaging device. Eleventh means for calculating the average value, the standard deviation, and the environmental data with a preset first reference value; and A thirteenth means for proceeding to a seventeenth means when the value is equal to or less than a value, and a case where the average value is larger than the first reference value and the standard deviation is equal to When the environmental data is equal to or more than the reference value, the fourteenth means proceeds to the seventeenth means, and when the standard deviation is smaller than the first reference value in the fourteenth means, If it is greater than or equal to the first reference value, the first
A fifteenth means for proceeding to the seventh means, a sixteenth means for judging that the environment data is in bad weather when the environmental data is smaller than the first reference value in the fifteenth means, The seventeenth means for judging that the weather condition exists.

Further, in the above-mentioned composite sensor type vehicle sensor according to the present invention, the eighth means may be arranged so that the infrared region image data acquired by the infrared region imaging device and the road surface data when no vehicle exists. A twenty-first means for comparing, a twenty-second means for proceeding to the twenty-sixth means when the temperature of the image data in the infrared region is equal to or lower than the temperature of the road surface data,
In the second means, when the temperature of the image data in the infrared region is higher than the temperature of the road surface data, a twenty-third means for calculating an area of the image data in the infrared region; Is smaller than the second reference value set in advance, the process proceeds to the twenty-sixth means.
Means for determining that the detected vehicle candidate is a vehicle when the area is larger than the second reference value in the twenty-fourth means; and The twenty-sixth means for determining that the object is an object.

[0014]

Embodiments of the present invention will be described below in detail with reference to the drawings.

[1] Description of Configuration FIG. 1 is a block diagram showing a basic configuration of a composite sensor type vehicle sensor according to an embodiment of the present invention. The composite sensor type vehicle sensor shown in FIG. 1 includes an imaging unit 10, an environment measurement unit 20, and a signal processing unit 30.
The imaging unit 10 includes a visible CCD camera 11 and an infrared imaging camera 12. Environmental measurement unit 2
0 has a configuration including an illuminometer 21 and a thermometer 22. The signal processing unit 30 includes a video A / D converter 31,
33; frame memories 32 and 34; a digital signal processor (hereinafter referred to as DSP) 35;
PU 36, external output I / F 37, A / D converter 3
8 and 39.

The image pickup unit 10 is installed above a road surface. The visible CCD camera 11 has a visible range (for example, 0.4 to 0.8 μm).
m) has sensitivity. The infrared region imaging camera 12 has sensitivity in the infrared region (for example, 3 to 5 μm or 8 to 12 μm).

The environment measuring section 20 is installed above the road surface integrally with the imaging section 10. The illuminometer 21 measures sunlight illuminance near the road surface. The thermometer 22 measures the outside air temperature near the road surface.

In the signal processing section 30, a video A / D converter 31 takes in the video signal of the visible CCD camera 11 and converts it into a digital image. The frame memory 32 is
The data of the video A / D converter 31 is held. Video A
The / D converter 33 takes in the video signal of the infrared camera 12 and converts it into a digital image. The frame memory 34 holds data of the video A / D converter 33.
The A / D converter 38 converts the signal of the illuminometer 21 into digital data. The A / D converter 39 converts the signal of the thermometer 22 into digital data. The digital signal processor 35 converts the image of the road surface and the vehicle in the visible region held in the frame memory 32 and the image of the road surface and the vehicle in the infrared region held in the frame memory 34 by the A / D converter 38. Based on the obtained illuminance data and the temperature data converted by the A / D converter 39, the detection of the vehicle, the distinction between the vehicle and the shadow of the building or the like and the vehicle, and the judgment of the weather condition are calculated. I do. The CPU 36 controls each unit and totals the detection results. The external output I / F 37 outputs the result totalized by the CPU 36 to the outside.

[2] Description of Operation FIG. 2 is a main flowchart showing the operation of the composite sensor type vehicle sensor according to one embodiment of the present invention. This shows a process for distinguishing a shadow from a shadow. FIG. 3 is a flowchart showing the weather state determination process in FIG. 2, and FIG. 4 is a flowchart showing the infrared region vehicle determination process in FIG. FIG. 5 is a diagram illustrating signal strengths of a visible image and an infrared image under each condition of the vehicle determination process.

First, a vehicle detection process in the configuration shown in FIG. 1 will be described with reference to FIGS. The illuminance meter 21 and the thermometer 22 are used to measure the illuminance and temperature near the road at regular intervals, and the data is obtained (F1). Visible region CCD camera 11 and infrared region imaging camera 1
2, a region including a preset road is imaged, and a two-dimensional image is obtained from each of the images (F2, F2
3). Based on the image data (hereinafter, referred to as a visible image) acquired by the visible region CCD camera 11, the temperature data and the illuminance data, the surrounding environment of the road surface is detected to determine the weather condition (F4). It is determined whether the vehicle determination is performed based on both the visible image data and the image data acquired by the infrared region imaging camera 12 (hereinafter, referred to as an infrared image) or only the infrared image data depending on the weather condition (F5). . If it is determined that the weather is normal, a vehicle candidate is detected using the visible image data. At this time, in addition to the vehicle, the shadow of the vehicle, the shadow of the building, and the like are all detected (F6). It is determined whether or not a vehicle candidate has been detected based on the visible image data. If a vehicle candidate has been detected, the process proceeds to detection of a vehicle candidate based on the next infrared image data. If no vehicle candidate is detected, the process returns to step F1 and repeats from F1 (F7). In both the case where it is determined that the weather is normal and the case where it is determined that the weather is bad, vehicle candidates are detected using infrared image data (F8). It is determined whether or not a vehicle candidate is detected based on the infrared image data. If a vehicle candidate is detected, a final vehicle determination is performed based on the results of and. If no vehicle candidate is detected, the process returns to F1 and repeats from F1 (F
9). If it is determined in F9 that the vehicle is a vehicle, the sensing output, speed, vehicle type, and the like are calculated and output as a final determination result (F10).

Referring to FIG. 3, a description will be given of the processing for determining the weather condition in F4 described above. In the visible image data, the average value and the variance value of the brightness data of the road surface in the preset area are calculated (F11), and compared with the known data (threshold) (F12). The average value of the luminance data is compared with the known data (F13). If the average value of the luminance data is equal to or smaller than the known data, it is determined that the weather is normal weather (F17). If the average value of the luminance data is larger than the known data, the determination is made using the variance value and the temperature data. If the variance is smaller than the known data (F14) and the temperature data is lower than the known data (F15), it is determined that the weather is bad (snowfall) (F16). Otherwise, if the variance is equal to or greater than the known data or the temperature data is equal to or greater than the known data, it is determined that the weather is normal (F17).

Referring to FIG. 4, the above-described vehicle determination processing in F8 will be described in detail. The same region as the position where the vehicle candidate was present in F6 is also processed for the infrared image data. However, when it is determined in F5 that the weather is bad, the area is not limited.

The currently acquired infrared image data is compared with the previously acquired background data (road surface data when there is no vehicle) (F21), and is the acquired data hotter than the road surface? Is determined (F2
2). If it is determined that the acquired data is hotter than the road surface, the area of the high-temperature area is calculated (F23), and the area is calculated based on the existing data (threshold).
It is determined whether it is larger than (F24). If the area is larger than the existing data (threshold),
It is determined that the vehicle is a vehicle (F25). In F22, if the acquired data is the temperature below the road surface or the area is less than the existing data, it is determined that the object is an object other than the vehicle (F26).

Next, a vehicle and an object other than the vehicle will be described with reference to FIG. 5 as an example of the signal strength of one line across the road in the visible image and the infrared image under each condition of the vehicle determination process.

FIG. 5A shows a case where a white vehicle exists in the first lane of a two-lane road and a black vehicle exists in the second lane. In the visible image data, the white vehicle has higher signal intensity (luminance) than the road surface, and the black vehicle has lower signal intensity (luminance) than the road surface. Further, in the infrared image data, both vehicles have a higher signal intensity (temperature) than the road surface due to heat of the vehicle body.

FIG. 5B shows that a vehicle exists in the first lane.
The case where the second lane has a shadow of the vehicle existing in the first lane is shown. In the visible image data, it is detected that a vehicle is present in the first lane, but in the second lane, the shadow may be erroneously detected as a black vehicle and processed as a vehicle candidate. In the infrared image data, if the shadow of the second lane is determined based on the temperature of the road surface, the shadow is not determined to be a vehicle, and only the vehicle in the first lane is detected.

FIG. 5C shows a case where a building shadow exists in both the first lane and the second lane. In the visible image data, a black vehicle may be erroneously detected as being present in both lanes or either lane and processed as a vehicle candidate. In the infrared image data, since the determination is based on the road surface temperature, it is determined that the vehicle is not a vehicle.

FIG. 5D shows a case where vehicles are present in the first lane and the second lane in bad weather. When the visible image has a low contrast such as in bad weather, the visible image data hardly detects the vehicle.
However, the vehicle can be detected based on the infrared image data.

FIG. 5E shows a case where vehicles exist in the first lane and the second lane at the time of snowfall. During snowfall, the road surface brightness level is higher than during normal weather. For this reason, in the visible image data, although the level slightly changes when the vehicle runs, it is difficult to detect the vehicle due to snow accumulated on the vehicle body. In the infrared image data, the heat of the hood or the window glass portion has a higher signal intensity than the road surface at the time of snowfall, and the vehicle can be detected.

[0030]

The first effect is to distinguish a vehicle from a road surface, a building, and the like by removing the influence of the shadow of the vehicle and the building, the road surface condition, and the surrounding environment of an object other than the vehicle. It is possible. This makes it possible to distinguish between an object other than a vehicle and a running or stopped vehicle without being affected by the surrounding environment, thereby improving the vehicle detection accuracy.

The reason is that a visible region imaging device having sensitivity in the visible light region, an infrared region imaging device having sensitivity in the thermal infrared region, and an environment measuring device for measuring the surrounding environment are used.
This is because the vehicle is detected from both the luminance, which is the signal intensity of the visible region image, and the temperature, which is the signal intensity of the infrared region image.

The second effect is that even in bad weather, the infrared region imaging apparatus which is not affected by changes in the surrounding environment of the vehicle can determine the detection of the vehicle.

The reason is that it is determined whether or not the weather is bad (such as snowfall) based on the average and variance values of the road surface brightness level and the temperature data by the visible region imaging device. The vehicle is detected using the imaging device.In bad weather, the vehicle detection result is inaccurate with the visible region imaging device.Therefore, using only the infrared region imaging device, the hood and window of the vehicle body are hotter than the road surface. This is because the vehicle is determined by detecting the glass and detecting the vehicle.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a basic configuration of a composite sensor type vehicle sensor according to an embodiment of the present invention.

FIG. 2 is a main flowchart showing an operation of the composite sensor type vehicle detector according to one embodiment of the present invention;

FIG. 3 is a flowchart showing a weather condition determination process in FIG. 2;

FIG. 4 is a flowchart showing processing for determining a vehicle in an infrared region in FIG. 2;

FIG. 5 is a diagram showing signal intensities of a visible image and an infrared image under each condition of a vehicle determination process.

FIG. 6 is a block diagram illustrating a configuration of an obstacle detecting device disclosed in Japanese Patent Application Laid-Open No. 5-296766.

FIG. 7 is a flowchart showing an image-based vehicle sensing method described in Japanese Patent Application Laid-Open No. 5-298594.

[Explanation of symbols]

 Reference Signs List 10 imaging unit 11 visible CCD camera 12 infrared region imaging camera 20 environment measuring unit 21 illuminometer 22 thermometer 30 signal processing unit 31, 33 video A / D converter 32, 34 frame memory 35 digital signal processor (DSP) 36 CPU 37 External output I / F 38, 39 A / D converter

Claims (3)

(57) [Claims] 1. A vehicle sensor for detecting a vehicle by imaging a road surface, wherein the vehicle sensor has sensitivity in a visible light region and has a visible region imaging device for imaging a visible region image, and has a sensitivity in a thermal infrared region. An infrared region imaging device that captures an image in the infrared region, an environment measurement device that includes an illuminometer and a thermometer, and measures an environment near a road surface; a visible region image data output by the visible region imaging device; A signal processing device that processes the image data in the infrared region output by the area imaging device and the environment data output by the environment measurement device to detect a vehicle, and the environment measurement device converts the environment data near the road surface to a predetermined value.
First means for acquiring at every time, and second means for acquiring an image of the visible region by the visible region imaging device.
Means, and the infrared region imaging device captures an image of the infrared region
Third means, wherein the signal processing device includes the environmental data and the image data in the visible region.
A fourth means for detecting the surrounding environment of the road surface using the data, and a sky environment based on the surrounding environment detected by the fourth means .
Fifth means for determining the weather condition, and the fifth means for determining that the vehicle is in a normal weather condition
In this case, the first image of the vehicle is
Means for detecting a candidate for the vehicle, and whether the first means for the vehicle has been detected by the sixth means .
If the first candidate for the vehicle is not detected
A seventh means for returning to the first means if the first candidate for the vehicle is detected by the seventh means.
Or the fifth means determines that the weather is bad.
In the case where the vehicle has been
Eighth means for detecting a second candidate of the vehicle, and whether the eighth means has detected a second candidate for the vehicle.
If no second candidate for the vehicle is detected,
Ninth means returning to the first means in the case where the second candidate of the vehicle is detected by the ninth means.
In this case, the final vehicle detection is performed and the detection result is output.
And a tenth means . 2. An eleventh means for calculating an average value and a standard deviation of road surface luminance data using image data of the visible region acquired by the visible region imaging device, A twelfth means for comparing the value, the standard deviation and the environmental data with a preset first reference value, and a seventeenth means when the average value is equal to or less than the first reference value. A thirteenth means for proceeding; and in the thirteenth means, when the average value is larger than the first reference value and the standard deviation is equal to or more than the first reference value, the seventeenth means. A fourteenth means for proceeding to the means, and if the standard deviation is smaller than the first reference value and the environmental data is equal to or greater than the first reference value in the fourteenth means, Fifteenth means for proceeding to the seventeenth means; and If the data is smaller than the first reference value, a first determination is made that the weather is bad.
The combined sensor type vehicle sensor according to claim 1, further comprising: means ( 6 ); and the seventeenth means for determining that the weather is in a normal weather condition. 3. The infrared means according to claim 8, wherein said eighth means compares the image data of said infrared region acquired by said infrared region imaging device with road surface data when no vehicle is present. If the temperature of the image data is equal to or lower than the temperature of the road data, a twenty-second means for proceeding to a twenty-sixth means, wherein the temperature of the image data in the infrared region is higher than the temperature of the road data in the twenty-second means If it is larger, a twenty-third means for calculating the area of the image data in the infrared region, and if the area is equal to or smaller than a second reference value set in advance in the twenty-third means, the twenty-sixth means is used. A twenty-fourth means for proceeding to the second means, and a twenty-fourth means for determining that the detected vehicle candidate is a vehicle when the area is larger than the second reference value in the twenty-fourth means. Vehicle candidate is something other than vehicle The composite sensor-type vehicle sensor according to claim 1 or 2 , further comprising: the twenty-sixth means for determining a body.