JP3257835B2 - Image type vehicle sensing device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image type vehicle which senses a vehicle running on a road or the like from a change in brightness level corresponding to a predetermined point on the road surface based on a video signal obtained by imaging the road or the like. It relates to a sensing device.

[0002]

2. Description of the Related Art In order to detect a vehicle traveling on a road,
A device for performing predetermined image processing based on a video signal obtained by capturing an image of a road with a TV camera to identify and detect a vehicle within an angle of view has been put to practical use.
No. 188005.

According to this proposal, it is possible to separately detect a vehicle traveling on a plurality of lanes (up to four lanes) with one image processing circuit corresponding to one TV camera on a single traveling road. is there. The apparatus and processing steps for this are described in detail in the above-mentioned Japanese Patent Application Laid-Open No. 4-188005, and are similar to the embodiment of the present invention described later. Therefore, here, a block diagram of an apparatus equivalent to FIG. The description is omitted. Note that the same reference numerals are given to the same parts as those of the apparatus of the embodiment.

[0004] By the way, a conventional image-type vehicle sensing device includes:
There is no particular inconvenience when targeting a single location,
In the case of installation at a plurality of locations, or when a large area such as a large intersection as shown in FIG. 9 is required and a plurality of sets of vehicle sensing devices are required, the appropriate number of vehicles (TV camera and processing circuit unit) are naturally used. And the number of vehicle sensing devices).

[0005] For example, if it is attempted to measure and monitor all of the traffic volume (total inflow traffic and all outflow traffic) at an intersection CRS in four directions with three lanes per direction shown in FIG. When a vehicle sensing device is used, eight or more conventional devices are required due to the limitation of the resolution.
The figure shows an arrangement of eight TV cameras CA1 to CA8. Eight processing circuit units (not shown) are required for each TV change, and there is a problem that the scale of the equipment becomes large and the price becomes difficult to actually introduce. Therefore, while waiting for efforts to reduce the price of the device, there has been a demand for an image-type vehicle sensing device capable of monitoring a wide range of traffic flow at a low price based on a new idea.

[0006] In addition to this problem, the conventional image-type vehicle sensing device detects headlights or small lights in order to detect a vehicle even at night, but this requires 30 lights per lane. Image processing must be performed by arranging ~ 50 measurement points. Therefore,
There was a problem that the monitoring area could not be so large.

[Problems to be solved by the invention]

[0007] Image vehicle sensing device according to the present onset Ming, above
Of which was made in view of the problem points, while intensive common technology and conventional image type automobile sensing device installed many have already been established, to ten to 12 units of about TV camera Image processing circuit of the same size as the conventional
By combining only the units, while suppressing the equipment scale,
An object is to propose a new device capable of monitoring a wide range of traffic flows. Incidentally, the price of a general image processing circuit is a price for several TV cameras. Therefore, sharing the image processing circuit is a very large difference in price.

Further, <br/> image vehicle sensing device according to the present onset Ming has been made to solve the other problems of the prior above, we propose a wide take apparatus target area at night surveillance.

Now , it is necessary to connect eight image cameras to one image processing circuit and measure the traffic flow of all the inflows and outflows at the intersection shown in FIG. 9 by the conventional image processing. If it is possible (that is, if the number of measurement points is set to be smaller per lane), the current image processing speed can be used. However, to cope with nighttime, 50 points × 8 (inflow / outflow direction) × 4 (number of lanes) ) = 1600 points, and luminance data must be obtained from an extremely large number of measurement points and processed. Therefore, it is very difficult to perform the same image processing as in the related art by one device due to the lack of speed of the arithmetic unit. To address this issue,
Even at night in other image vehicle sensing device according to the present onset bright, the image processing circuit of the conventional level of Speed simply only combine one against ten to 12 units of about TV camera,
It is an object of the present invention to propose an image type vehicle sensing device capable of processing a large amount of traffic in real time.

[0010]

Image vehicle sensing device SUMMARY OF THE INVENTION The first invention comprises: a synchronizing signal generating circuit for generating an external synchronizing signal, the same external synchronization generated by the synchronization signal generating circuit
Signal driven and overlapping measurement lines
A plurality of TV cameras set on scanning lines different from each other and individual video signals from these TV cameras are input, and these video signals are set to the measurement line.
Scanning lines are sequentially switched so that scanning lines are included, and scanning is performed.
Image processing for detecting a synchronization selection circuit for outputting the synthesized composite video signal for each査線, the vehicles existing video signal to the image processing to the measurement lines of each TV camera from its synchronous selection circuit (20) And a part.

Further, the image type vehicle sensing device according to the second invention of the present application comprises: a synchronization signal generating circuit for generating an external synchronization signal;
By the same external synchronization signal generated by the synchronization signal generation circuit
As well as the measurement lines
Multiple TV cameras set on different scanning lines
Individual video signals from these TV cameras are input and
These video signals are included in the scanning line on which the measurement line is set.
Switching in units of scanning lines so that it is rare,
A synchronization selection circuit for outputting a synthesized video signal;
Image processing of the video signal from the synchronization selection circuit of
The average brightness level is obtained, and a predetermined value is deviated from this average brightness level.
Binarized output comparing only high level signal and video signal
, And perform image processing based on the binarized output to
Image processing unit that detects vehicles existing in the field of view at night
And is configured.

Further, the image type vehicle sensing device according to the third invention of the present application is a synchronous signal generating circuit for generating an external synchronous signal, and is driven by the same external synchronous signal generated by the synchronous signal generating circuit, and has a measuring line. A plurality of TV cameras set on mutually different scanning lines without overlapping, and individual video signals from these TV cameras are input so that these video signals are included in the scanning lines on which the measurement lines are set. A synchronous selection circuit for sequentially switching in units of scanning lines and outputting a composite video signal synthesized for each scanning line, and performing image processing on the video signal from the synchronous selection circuit to set the average luminance level of the road surface to 5 points per lane 1 point below
Obtained from the luminance data at the above measurement points, a level signal higher than the average luminance level by a predetermined value and a video signal are compared to obtain a binarized output, and image processing is performed based on the binarized output. And an image processing unit for detecting a vehicle existing within the field of view of each TV camera at night .

[0013]

In the device according to the first aspect of the present invention, the synchronization signal generating circuit
Operate at the same timing with the same external synchronization signal.
Video signals from a plurality of TV cameras
Input and synchronized with the synchronization signal1Switching sequentially for each scanning line
Output. The composite video signal output in this way is
Is processed by a single image processing unit in almost the same way as before.
From the image of one frame for multiple TV cameras,
Therefore, vehicles on a plurality of road surfaces are detected.

In the apparatus of the second invention of the present application, only one
Image processing unit, as in the first aspect of the present invention,
Processed and multiple TV cameras from one frame image
Min, and therefore target vehicles on multiple roads
Next, image processing of the video signal from the TV camera is performed to
The average brightness level is obtained, and a predetermined value is deviated from this average brightness level.
Generates a high analog level signal,
Is always compared to obtain a binary output, and this output is
Perform image processing in association with each of multiple TV cameras at night.
Detect vehicles present in the field of view.

Further, in the device according to the third aspect of the present invention, only one image processing unit processes the combined video signal in the same manner as in the first aspect of the invention, and converts a single frame of image into a plurality of TV cameras, and thus a plurality of TV cameras. Targets vehicles on the road surface and performs image processing on the video signal from the TV camera to reduce the average luminance level of the road surface to 5 points or less per lane.
Obtained from the luminance data at the above measurement points, an analog level signal higher than the average luminance level by a predetermined value is generated, and this is constantly compared with the video signal to obtain a binary output. The image processing is performed in association with the image, and a vehicle existing in each field of view of a plurality of TV cameras at night is detected .

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Each invention of the present application will be described below with reference to the drawings based on embodiments. FIG. 1 is a block diagram showing an image type vehicle sensing device according to an embodiment of the present invention. The image-type vehicle sensing device of the embodiment can be roughly divided into a plurality (eight units).
TV camera CA (CA1 to CA8), a synchronizing signal generating circuit 10 for supplying the same synchronizing signal thereto, a synchronizing selecting unit 20 for sequentially selecting and outputting output video signals of each TV camera CA based on the synchronizing signal, And an image processing unit 30 that performs image processing on the selectively output video signal to detect vehicles existing within the field of view of each TV camera.

For the purpose of detecting a vehicle at night, the image processing unit 30 in the apparatus of the embodiment uses a level higher than the average luminance level of the road surface obtained by performing image processing on the video signal from each TV camera by a predetermined value. A video comparison circuit 40 for comparing a signal and a video signal in an analog manner and outputting the result is included. Hereinafter, each part will be described in detail.

The TV camera CA is, for example, a CCD camera and is driven by the same external synchronization signal.
As indicated by A8, for example, two cameras are provided at each of the four corners of the target intersection CRS, for example, eight in total, and images of the detection area of the intersection are taken from above, and the corresponding NTSC video signal is transmitted to the synchronization selection unit 20. Synchronous signal generation circuit 10
Generates the external synchronizing signal, supplies it to each TV camera CA, and outputs the same synchronizing signal to a synchronizing signal extracting circuit 121 included in the image processing unit 30 described later in detail.

The synchronization selection section 20 is composed of an input selection circuit 21 and a video signal switching circuit 22, and switches the input (at a fixed timing) based on the synchronization signal. Video signals from the plurality of TV cameras CA are individually input to the video signal switching circuit 22. Also, the input selection circuit 21
Includes a microprocessor 141 included in the image processing unit 30.
The camera selection signal associated with the synchronization signal is also input, and controls the video signal switching circuit 22 in accordance with the content of the camera selection signal. In response to this, the video signal switching circuit 22 selects any one of a plurality of video signals from each TV camera, and from its output terminal, a video signal input terminal of the image processing unit 30 (video amplifier 127).
Let through.

The image processing unit 30 controls the entire apparatus and performs predetermined image processing. The image processing unit 30 generally includes an arithmetic control unit including a luminance data conversion unit 100 and a microprocessor 141.
Consists of 200. Note that the luminance data conversion unit 100
Includes a video comparison circuit 40.

The luminance data conversion unit 100 has a synchronization signal extraction circuit 121 connected to the synchronization signal generation circuit 10 for obtaining a horizontal synchronization signal and a vertical synchronization signal. Based on both synchronization signals, a current search position on a search line is determined. Horizontal address counter 12 for digitizing and obtaining corresponding horizontal and vertical coordinates
2 and a vertical address counter 123, a digital comparator 124 for comparing these outputs with the coordinates designated by the microprocessor 141 in sequence through an input interface 129 and a preset counter 126, and generating an acquisition signal at the moment of coincidence, A video A / D conversion circuit for quantifying the luminance at the corresponding instant of the video signal from the synchronization selecting section 20 in accordance with the fetched signal and obtaining and outputting luminance data
125 is included. The input part has a video amplifier 127 that amplifies the transmitted video signal to an appropriate level.
Following this, a clamp level fixing circuit 128 for normalizing the video signal to a constant amplitude prior to the A / D conversion is also provided.

A video comparison circuit 40 included in the luminance data conversion section 100 is for detecting a vehicle, especially at night. The image processing section 30 converts a video signal from each TV camera into an image as described later. An analog level signal (threshold value Ts) higher by a predetermined value (Lc) than the average luminance level La of the road surface obtained by processing and a video signal (obtained from the clamp level fixing circuit 128 in the embodiment) are input. Then, the two are compared in an analog manner and the result is output. The video comparison circuit 40 is connected to the D / A converter 4 connected to the CPU bus 145 of the microprocessor 141.
1 and the video comparator 42 whose output is connected to one input terminal Ia and the output (video signal) of the clamp level fixing circuit 128 is connected to the other input terminal Ib, and a memory 43 subsequent thereto. It is configured. Memory 43
Stores the comparison result (binary data) of the video comparator memory 42 and outputs the result to the CPU bus 145 upon request.

The arithmetic control unit 200 controls the above-mentioned units and performs high-speed arithmetic processing such as averaging a plurality of luminance data and generating a correlation upon receiving the luminance data from the video A / D conversion circuit 125. After that, the obtained data is further subjected to predetermined processing to detect the presence of a vehicle, calculate the speed of the vehicle, and output the result to a subsequent device.

The arithmetic control unit 200 requires a high-speed operation for averaging the above-mentioned luminance data and calculating the correlation, so that the signal processor 131 (including a storage unit necessary for storing data) is provided. ), A microprocessor 141, a ROM 142 storing a control program,
A RAM 143 for data storage and an I / O circuit 144 for output are provided. Reference numeral 145 denotes a CPU bus through which the microprocessor 141 exchanges data and addresses, control commands and response signals with other parts.

An adjustment tool 300 (a block diagram is shown in FIG. 2) is connected to the arithmetic control unit 200 via an input interface 146. The image processing unit 30
Although not shown, there are also a clock circuit supplied to each section, a D / A conversion circuit for sending out a signal for increasing the sensitivity of the TV camera when the level of the video signal from the TV camera is low and inappropriate for processing. Etc. are provided.

FIG. 2 shows the image processing unit 30 shown in FIG.
Adjustment tool 300 and monitor TV required for device setting
400 and a printer 500. These are auxiliary devices that are required only when setting the position of each measurement point so that the image-type vehicle sensing device is suitable for each actually installed site. The detailed configuration and usage of these auxiliary equipment are described in detail in the above-mentioned specification of Japanese Patent Application Laid-Open No. Hei 4-188005.

Next, the operation of each part of the apparatus of the embodiment will be described (see FIGS. 1 and 2). Multiple TV cameras CA1
To CA8 are installed at the corners of the intersection as shown in FIG.
The lane is imaged from above, and corresponding video signals are obtained. FIG. 3 shows an image PIC corresponding to such a video signal.
1 to PIC8 are schematically shown. In addition, CL1 in each image
To CL8 are scanning lines VA1 to VA at predetermined positions (different from each other) whose scanning directions substantially coincide with the cross direction of the monitored road.
A plurality of measurement points are set along each measurement line on the measurement line set on 8. Each TV camera C
A performs an imaging operation in synchronization with an external synchronization signal supplied from the synchronization signal generation circuit 10, and corresponding points on each screen (for example, points PT1 to PT8 at the upper left of the screen) are obtained at the same moment. , For example, the last scanning line VA of one frame
Z is also scanned simultaneously by eight units.

Each video signal is input to the synchronization selector 20. The synchronization selection unit 20 sequentially switches the plurality of input video signals at a high speed in a predetermined order in synchronization with the external synchronization signal and outputs the signals to the image processing unit 30. In the embodiment, the video signals from the plurality of TV cameras are switched by the video signal switching circuit 22 in units of one scanning line corresponding to the time of the designated scanning line number. That is, a signal obtained from the synchronization signal extraction circuit 121 via the digital comparator 124 and a signal from the microprocessor 141 via the CPU bus 145 are input to the camera selection circuit 23 from the signal from the synchronization signal generation circuit 10. Have been.

The camera selection signal generating circuit 23 sequentially sends out camera selection signals for each scanning line to the input selection circuit 21 based on both inputs to the input selection circuit 21. The input selection circuit 21 responds to the camera selection signals by using the TV cameras CA1 and CA1. A signal for selecting CA2,..., CA8 in this order is generated (decoded) and output to the video signal switching circuit 22. Therefore, the video signal switching circuit 22
Output, a so-called synthesized video signal obtained by synthesizing the video signal from each TV camera for each scanning line is obtained.
Output to 0. FIG. 3 shows a (synthesized) image PIC 'corresponding to such a synthesized video signal.

FIGS. 4 and 5 are time charts for explaining this process. The current video switching device has a response frequency of 1 with a switching time of 0.5 microsecond or less.
There is no problem because a high-speed one of 0 MHz or more can be easily obtained. The camera selection signal is output with a delay of time Td from the horizontal synchronization signal as shown in FIG. In this way, the blanking period is avoided and the horizontal synchronizing signal is not erased.

As shown in FIG. 6, a BCD-decimal decoder is used for the input selection circuit 21 for camera selection, and unless a new selection is made, the video signal of the previous TV camera is continuously selected. It is in.

After the synthesized video signal is obtained once, as described later, conventional image processing (for example,
The vehicle detection for eight TV cameras can be performed by image processing substantially similar to that described in Japanese Patent Application Laid-Open No. 188005). Each image is simply taken into account by taking into account additional information on which TV camera corresponds to the obtained measurement line. Processing may be performed. That is, signals from a plurality of TV cameras are simultaneously processed by one image processing unit, and traffic flow measurement at a plurality of locations can be performed.

In the embodiment, eight video signals are switched for each scanning line. However, one scanning line can be further subdivided to synthesize a plurality of video signals. For example, the first half of one scanning cycle is a TV camera. CA1 video signal and TV camera C in the latter half
The A2 video signal may be a composite video signal, or the TV camera CA1 and the TV may be connected at a period of, for example, 1/128 of one scanning cycle.
It is also conceivable to alternately switch the video signal of the camera CA2 to obtain a composite video signal.

The process of processing a (synthesized) video signal in the present application will be described below. The image processing unit 30 performs A / D conversion of the luminance level of the input (synthesized) video signal to perform digital image processing. In other words, similar to general image processing,
First, as shown in FIG. 3, using the video signal, the measurement lines CL1 to CL are displayed on the video screen of each camera by using an adjustment tool.
8 and a number of measurement points on each measurement line. At this time, the respective addresses (vertical scanning line numbers) VA1 to VA8 of the measurement lines CL1 to CL8 must not overlap each other. Note that this setting is made in Japanese Patent Laid-Open No. 4-188.
As in the case of No. 005, the personal computer of the adjustment tool 300 is operated and set on the Freeze image in advance, and is stored in the image processing unit 30.

[0035] At the time of operation, field luminance data group in which the luminance <br/> level in the luminance signal temporally corresponding to said plurality of measurement points set in the image and each determined pair < br /> sequentially determined for each de sufficient difference observed depending on the comparison between when the vehicle is not present brightness data group consisting of the luminance data of each measurement point (road surface only corresponds to) (reference road surface level data) The obtained data is regarded as the vehicle corresponding point, and the existence of the vehicle at each measurement point at a certain time is detected. However, in the embodiment, since images of different TV cameras, that is, data of different areas are processed for each scanning line, the vehicle sensing process needs to be performed for each scanning line.

For detecting a vehicle, the luminance data converter 120 first obtains a horizontal synchronizing signal and a vertical synchronizing signal from the synchronizing signal from the synchronizing signal generating circuit 10 by the synchronizing signal extracting circuit 121. The horizontal address counter 22 and the vertical address counter 23 operate based on these synchronization signals,
The current search position on the search line is shown one after another. In addition,
These horizontal synchronizing signal and vertical synchronizing signal are denoted by reference characters A in FIG.
It may be separated from the output video signal (via a video amplifier) from the video signal switching circuit 22 by the connection of the dashed line with.

The output indicating the operation position is compared with each coordinate designated by the microprocessor 141, and the digital comparator 124 generates a capture signal at the moment of coincidence. In response to the fetched signal, the video A / D conversion circuit 125 quantifies the luminance at the corresponding instant of the video signal and outputs it as luminance data along with the corresponding horizontal and vertical coordinates.
In the input section, the transmitted video signal is amplified to an appropriate level by a video amplifier 127, and a subsequent clamp level fixing circuit 128 normalizes the video signal to a constant amplitude prior to A / D conversion.

In this manner, the luminance level at the temporal position in the luminance signal corresponding to each position of a plurality of measurement points set in substantially the entire area of the image PIC 'in FIG. br /> bell is obtained as a digital value. These brightness levels set in which the luminance data group or al 1 screen (1 frame) is binarized vehicle binarized data groups are obtained. That, <br/> luminance data groups consisting of luminance data at each measurement point when the vehicle is constantly updated at appropriate intervals absent (corresponding only to the road surface) (reference road surface data) and luminance data groups The presence of a vehicle at each measurement point at a certain time is detected as a vehicle corresponding point using data for which a sufficient difference is recognized as a result of the comparison.

For this purpose, the signal processor 131 receives the luminance data from the video A / D conversion circuit 125, averages a plurality of luminance data at high speed, and calculates a correlation, and then at appropriate time intervals. extracted is compared with a reference road data comprised of the luminance level of the road surface, binarizes the luminance data for each measurement point, the binary data groups a set of corresponding to one field is determined. In addition, each of the above processes,
This is a technique similar to the conventional image-type vehicle detection method. For example, Japanese Patent Application Laid-Open No. 4-188005 mentioned above also details vehicle detection by pattern correlation comparison between a road luminance pattern and a vehicle luminance pattern.

The microprocessor (host CPU) 141 of the arithmetic and control unit 40 stores the vehicle binary data for each frame.
And the RAM 143 sequentially stores the vehicle binarized data corresponding to each address. Microprocessor
141 controls the other units and receives the vehicle binarization data , and scans the data on a scan line basis (that is, for each area) based on the data.
Performs predetermined processing, detects the presence of vehicles for each target area,
A vehicle detection signal is transmitted to a predetermined location through the I / O circuit 141, and a predetermined display is performed.

[0041] As described above, the image type according to the present onset Akira
In the vehicle sensing device , a plurality of TV cameras driven by an external synchronization signal, a synchronization signal generation circuit, and each of the T cameras during one frame period based on the synchronization signal.
It comprises a synchronization selection circuit for sequentially switching and outputting each video signal from the V camera, and an image processing unit for performing image processing on the video signal from the synchronization selection circuit to detect vehicles in a plurality of monitoring areas.

Next, a description will be given of another image vehicle sensing device according to the present onset bright. The background of these inventions will be described again very simply. Including the above-mentioned images vehicle sensing device,
In general, an image-type vehicle sensing device does not work at nighttime because the reflected light from the vehicle to be detected is reduced, as in the daytime. Of course,
There is also a solution of irradiating sufficient illumination light, but it is not often used from the viewpoint of power saving and cost reduction.

Instead, since the vehicle to be detected lights the headlights or the like at night, it is common to detect the vehicle by processing the headlight-corresponding portion in the video. In such nighttime headlight detection, since the bright spot is small, it is necessary to set 30 to 50 measurement points per lane. That is, more information must be processed than in the daytime, and an extremely large amount of information processing is necessary to obtain video signals from eight or more TV cameras and detect headlight signals as in the present application. The required processing cannot be performed at the current processing speed of the device.

Note that even in a conventional apparatus to which one camera is connected, three cameras per lane are used at night (headlight).
Assigning 0 to 50 points inevitably reduces the monitoring area. Other image-type vehicle according to the present onset Akira
The sensing device is provided to solve this inconvenience. The set measurement points per lane are the same as those in the daytime, for example, 10 points.
Even if the number of points is small, it is possible to obtain an effect equivalent to performing image processing of 30 to 50 points per lane and detect headlights at night and small lights at twilight to detect a vehicle. It is an invention.

For this reason, in the previous embodiment shown in FIG. 1, the image processing section 30 performs image processing on the video signal from the TV camera to obtain an average luminance level of the road surface, and a predetermined value from the average luminance level. A high-level signal and a video signal are compared to obtain a binarized output, and image processing is performed based on the binarized output so that a night vehicle (headlight) can be detected.

To this end, the image processing section 30 includes a video signal comparing section 40. Video signal comparison unit 40
Is a D / A converter 41 for D / A converting the output from the microprocessor 141, as shown in the block diagram of FIG.
And the video comparator 4 connected to the D / A converter 41.
2, and a memory 43 subsequent thereto.
The video comparator 42 is a two-input analog comparator. One input terminal Ia is connected to the output of the D / A converter 41 as described above, and the other input terminal Ib receives the (composite) video signal. Has been entered. In this embodiment, the output of the clamp level fixing circuit 128 is input as a video signal. The output of the digital comparator 124 is also connected to both the video comparator 42 and the subsequent memory 43. The memory 41 is connected to the CPU bus 145 so that the microprocessor 141 can read the storage result.

Hereinafter, the operation of the apparatus of the embodiment shown in FIG. 1 at night will be described. Each part obtains the average road luminance by performing the same operation as described above in the image type vehicle sensing device . That is, when briefly described again, the luminance level (Cij) in the luminance signal temporally corresponding to a plurality of measurement points set in the image is obtained, and from this, the luminance data group (Dt) is sequentially obtained for each field. And it is always updated at appropriate intervals, when there is no vehicle (only for road surface)
A luminance data group (reference road surface data, Dr) composed of luminance data at each measurement point is obtained by calculation.

[0048] The reference road surface data (average road luminance) is,
As in the daytime, image A / D conversion is performed based on measurement points of 5 points or less per lane to learn the average value. The value of the reference road surface data Dr thus obtained is input to the D / A converter 41 of FIG. In the D / A converter 41, as shown in FIG. 8, a detection threshold Ts higher than the average value La corresponding to the input value by a predetermined level Lc is determined, and based on this, the detection threshold Ts is compared with the input video signal Lt. Then, a headlight detection signal is obtained. That is, the headlight detection signal is output to the memory 43 only while the instantaneous value of the video signal Lt is higher than the threshold value.

The predetermined level Lc may be added by the D / A converter 41, but is calculated by the microprocessor 141 to obtain a digital value corresponding to the detection threshold Ts. Threshold T
s may be input to the D / A converter 41.

The memory 43 stores the headlight detection signal and outputs it in response to a read request from the microprocessor 141. Thus, the headlight detection signal is captured corresponding to the same measurement point as in the daytime. Subsequent processing is
The vehicle can be detected even at night by performing the above-described processing that is substantially the same as during the day. Since the number of measurement points to be processed is the same as in the daytime, (composite) video signals obtained from eight or more TV cameras by a predetermined rule can be processed.

[0051] As described above, another image vehicle sensing device according to the present onset Ming, image type according to the present onset bright vehicle sensing apparatus <br/> same, and a plurality of TV cameras, the synchronization signal generation circuit And a synchronization selection circuit, and further performs image processing on the video signal from the input selection circuit to obtain an average luminance level of the road surface, and compares the level signal with the video signal by a predetermined value higher than the average luminance level. An image processing unit that obtains a binarized output, performs image processing based on the binarized output, and detects a vehicle existing in the field of view of each TV camera at night.

[0052] Incidentally, another image vehicle sensing device according to the present onset Ming, which was used in combination the technical concept described above a single TV camera, in this case, finely set the measuring points at night Since there is no need to perform this, there is an effect that the vehicle can be detected in a wider range than before.

[0053]

Image vehicle sensing device according to the present onset bright as detailed above, according to the present invention, a plurality of TV that is driven by an external synchronization signal
A camera, a synchronization signal generation circuit common to them, and each T
Since the video signal from the V-camera is input and the video signal from the video synchronization selection circuit is output to one image processing unit since the video signal from the video synchronization selection circuit is sequentially switched and output during one frame period based on the video synchronization signal. Image processing with each T
Vehicles in the field of view of the V-camera, and thus a very wide range of monitored objects, can be handled. That is, the image-type vehicle sensing device for a wide area such as a large-scale intersection has an effect of being extremely inexpensive since there is only one image processing unit.

[0054] The image vehicle sensing device according to the present onset bright, with the average luminance level of the road surface by image processing a video signal from the TV camera, and only a high-level signal by a predetermined value amount than the average brightness level It is configured to include an image processing unit that obtains a binarized output by comparing with a video signal, performs image processing based on the binarized output, and detects a vehicle existing in the field of view of the TV camera at night. Since there is no need to finely set the measurement points even at night, there is an effect that the vehicle can be detected in a wider range than before.

[0055] Furthermore, the image type vehicle sensing device according to the present onset Ming, outside a plurality of TV cameras is driven by the synchronization signal, and the synchronizing signal generating circuit common to these, each of the video signals by each TV camera is input A synchronization selection circuit for sequentially switching and outputting each video signal during one frame period based on the synchronization signal; and performing image processing on the video signal from the TV camera to obtain an average luminance level of a road surface. A level signal higher than the average luminance level by a predetermined value and a video signal are compared to obtain a binarized output, and image processing is performed based on the binarized output to detect a vehicle existing in the field of view of the TV camera at night. It is not necessary to finely set measurement points even at night, so that the video signal from the synchronization selection circuit is image-processed by one image processing unit even at night. Vehicle that exists in the field of view of the V camera,
Therefore, an extremely wide range of monitored objects can be handled. That is, it is possible to realize an image-type vehicle sensing device that can be used at night and that is compatible with a wide area such as a very large-scale intersection.

[Brief description of the drawings]

FIG. 1 is an overall block diagram showing an embodiment of an image type vehicle sensing device according to each invention of the present application.

FIG. 2 is a block diagram of an image processing unit and an associated adjustment tool of the imaging vehicle sensing device of FIG. 1;

FIG. 3 is an explanatory diagram showing each image of the image-type vehicle sensing device according to the present application.

FIG. 4 is a time chart for explaining the operation of the image-type vehicle sensing device according to the present application.

FIG. 5 is a time chart illustrating the operation of the image-type vehicle sensing device according to the present application.

FIG. 6 is a circuit diagram illustrating an example of an input selection circuit according to the present application.

FIG. 7 is a circuit block diagram illustrating an example of a video comparison unit according to the present application.

FIG. 8 is a diagram illustrating an operation of a video comparison unit according to the present application.

FIG. 9 is a schematic plan view showing a large-scale intersection suitable for applying the image-type vehicle sensing device of the present application.

FIG. 10 is an overall block diagram showing an example of a conventional image-type vehicle sensing device.

[Explanation of symbols]

 CA (CA1 to CA8): TV camera, 10: synchronization signal generation circuit, 20: synchronization selection circuit, 30: image processing unit, 40: video comparison unit.

──────────────────────────────────────────────────続 き Continued on the front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) G08G 1/04 G06T 1/00

Claims (3)

(57) [Claims] A synchronous signal generating circuit for generating an external synchronizing signal; and a synchronous signal generating circuit which is driven by the same external synchronizing signal generated by the synchronizing signal generating circuit. Multiple TV cameras (CA) set to
The individual video signals from these TV cameras (CA) are input, and these video signals are sequentially switched in units of scanning lines so as to include the scanning lines in which the measurement lines are set, and are synthesized for each scanning line. Synchronization selection circuit that outputs a composite video signal
(20) and an image processing unit (30) configured to image-process a video signal from the synchronization selection circuit (20) to detect a vehicle existing in a measurement line of each TV camera. Sensing device. 2. A synchronizing signal generator for generating an external synchronizing signal.
Circuit (10) and the same external synchronization signal generated by the synchronization signal generation circuit.
And the measurement lines are not overlapped
Multiple TV cameras (CA) set on different scanning lines
And each video signal from these TV cameras (CA) is input.
These video signals are scanned with the measurement line set.
Scans are switched sequentially in scan line units so that lines are included.
Synchronous selection circuit that outputs a synthesized video signal synthesized for each line
(20) and the video signal from the synchronization selection circuit (20) are image-processed and routed.
Obtain the average brightness level of the
Binarize by comparing the level signal and video signal that are higher by the fixed value
And outputs an image based on the binarized output to obtain each T
Detects vehicles existing in the field of view of the V camera (CA) at night
An image-type vehicle sensing device including an image processing unit (30);
Place. 3. A synchronizing signal generator for generating an external synchronizing signal.
Circuit (10) and the same external synchronization signal generated by the synchronization signal generation circuit.
And the measurement lines are not overlapped
Multiple TV cameras (CA) set on different scanning lines
And each video signal from these TV cameras (CA) is input.
These video signals are scanned with the measurement line set.
Sequentially switches the scanning line basis to include a line scan
Synchronous selection circuit that outputs a synthesized video signal synthesized for each line
(20) and the video signal from the synchronization selection circuit (20) are image-processed and routed.
The average brightness level of the surface should be less than 5 points per lane
From the brightness data at measurement points above
Level signal and video signal higher than the luminance level by a predetermined value
To obtain a binarized output, and based on this binarized output
Vehicles that exist in the field of view of each TV camera (CA) after image processing
Image processing section for detecting the at night (30) and a configured image
Image type vehicle sensing device.