JPH04127280A - Tunnel detector - Google Patents

Abstract

PURPOSE:To perform fast discrimination and to obtain high discrimination accuracy by discriminating the case where the continuous black level area with prescribed width of the binary video signal of a one-dimensional image pickup device in which a picture element row is arranged in a lateral direction exists in the center part of an image pickup area, and the case where a pair of areas with prescribed width exist at both sides of the center part of the image pickup area as a tunnel. CONSTITUTION:This detector is equipped with the one-dimensional image pickup device 1, a comparator 2 as a binarization processing means, and a microcomputer 3 as a tunnel discrimination means, etc. In such a case, a binary signal Vd of one picture element row is received from the comparator 2 in an odd-numbered scanning period, and it is stored in a build-in storage area transiently. The binary signal Vd is processed in the next even- numbered scanning period, and all the continuous black level areas Rb are extracted, and the start points and the end points of the areas are stored. When the continuous black level area of the binary video signal with prescribed width exists in the center part of the image pickup area of the one-dimensional image pickup device, or when the pair of areas with prescribed width exist at both sides of the center part of the image pickup area of the one- dimensional image pickup device, it is discriminated as the tunnel. In such a way, the fast discrimination can be performed, and the high discrimination accuracy can be obtained in spite of simple device constitution.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a tunnel detection device. This device can be applied to an automatic light control device, an automatic outside air switching device for a car air conditioner, an interruption warning device for a car phone, etc.

[Prior Art] Conventional auto light control devices for vehicles operate based on changes in illuminance, so they cannot distinguish between tunnels and bridge girders, and have a problem in that they turn on when passing a bridge girder.

In order to eliminate this drawback, Japanese Patent Application Laid-Open No. 60-151521
Publication No. 63-7977 is based on sound waves (noise).
Accordingly, JP-A-54-158962 detects tunnels using self-emitted ultrasonic waves, and JP-A-60-240545 uses a two-dimensional image sensor to detect tunnels.

[Problems to be Solved by the Invention] The method of detecting tunnels using a two-dimensional imaging device has the advantage of being able to detect tunnels earlier than the acoustic method;
There are various patterns in the two-dimensional shape of tunnels, and a large-scale image processing device is required to distinguish them, and the time required for image processing cannot be ignored.

For example, the upper part of the tunnel may be empty, a vertical concrete surface, a tree surface, or a rock surface, and it is not easy to set the black level of the binarized imaging device to only the tunnel part. Noise is mixed in. Furthermore, it is not easy to accurately identify tunnel portions with various cross-sectional shapes using this binarized two-dimensional imaging device.

The present invention has been made in view of these problems, and it is an object of the present invention to provide a tunnel detection device that is capable of high-speed discrimination and can obtain high discrimination accuracy despite a simple device configuration.

[Means for Solving the Problems] The tunnel detection device of the present invention includes a one-dimensional image carrier that is mounted on a vehicle toward the vehicle traveling direction and visualizes the image in the lateral direction, and a first-dimensional image carrier output from the one-dimensional image pickup device. A binarization means for binarizing the original imaging device; and a tunnel discrimination means for processing the binarized imaging device outputted from the binarization means to discriminate a tunnel, and the tunnel discrimination means includes the When a continuous black level area in which the black level signal of the binarized image pickup device is continuous exists with a predetermined width in the center of the R image area of the -dimensional image transfer device, and It is characterized in that a tunnel detection signal is output when the tunnel detection signals are present in pairs with a predetermined width on both sides of a predetermined narrow width in the center of the image area.

The one-dimensional imaging device that covers the image in the horizontal direction may be any device that can obtain an imaging device with approximately one row in the horizontal direction, and it is also possible to use one pixel row of an area image sensor.

1 effect] When a one-dimensional imaging device obtained by visualizing a tunnel with a one-dimensional imaging device in which pixel rows are arranged in the horizontal direction is binarized by a binarization means, continuous black of the binarized imaging device is generated. The level areas exist with a predetermined width in the center of the image development area of a one-dimensional image transfer device, or exist in pairs with a predetermined width on both sides of a predetermined narrow width in the center of the image development area of a -dimensional image transfer device. do.

Therefore, when these two one-dimensional image bangs occur, it can be determined that it is a tunnel.

[Example] [First example] A tunnel detection device according to a first embodiment is shown in FIG.

This device includes a -dimensional image carrier 1, a comparator 2 as a binarization processing means, a microcomputer 3 as a tunnel discrimination means, and an autolight control device 4 driven by the microcomputer 3.

The one-dimensional image carrier 1 is mounted on the dash board of the vehicle as shown in FIG.
An OCD linear image sensor 1b and a lens system 1a are installed.The optical axis of the lens system 1a extends almost horizontally in the direction of vehicle travel, and the pixel rows of the CCD linear image sensor 1b extend in the direction of vehicle travel. They are arranged approximately horizontally.

The light incident through the windshield is imaged by the lens system 1a onto the pixel row of the CCD linear image sensor 1b,
The CCD linear image sensor 1b is a one-dimensional imaging device Vs
is output to comparator 2. Incidentally, the scanning period of the CCD linear image sensor 1b is 64.5 μsec.

The comparator 2 compares the one-dimensional imaging device Vs with a predetermined threshold voltage Vref, and converts the one-dimensional imaging device Vs into a binary-dimensional imaging device (hereinafter referred to as
It outputs Vd (simply referred to as a binary signal). The microcomputer 3 processes the binary signal Vd to determine a tunnel, and outputs a lighting signal to the auto light control device 4 when determining the tunnel.

Here, the threshold voltage Vref is set to a level that makes the output signal voltage of the pixel that visualizes the inside of the tunnel O.

Next, the operation of the microcomputer 3 will be explained with reference to the flowchart shown in FIG.

First, the binarized signal Vd for one pixel row is received from the comparator 2 during an odd scanning period, and is temporarily stored in a built-in storage area (S10).

In the next even scanning period, the binarized signal V accumulated for - hours
d to extract all continuous black level regions Rb,
The starting point and ending point are stored (S12). Note that the starting point coordinates of the continuous black level region Rb are the coordinates of a pixel where the pixel adjacent to the left side is at the white level and the pixel itself is at the black level, and the end point coordinates of the continuous black level region Rb are the coordinates of the pixel where the pixel adjacent to the left side is at the black level. is the coordinate of the pixel that itself is the white level,
The continuous black level region Rb means a pixel region from the above-mentioned starting point to the above-mentioned ending point.

Next, at least one of the central coordinates (intermediate coordinates between the start point and the end point) of each continuous black level region Rb is the entire pixel region 1ij
! Check whether it exists in the center of (linear) (314
), if the continuous black level region Rb exists in the center, it is checked whether the width of the continuous black level region Rb is equal to or larger than a predetermined value (here, half or more of the total pixel region width). If it is more than half of the total pixel area width, it is determined that it is a tunnel and the automatic light control device is turned on in step 322; otherwise, the process proceeds to step 328.

Further, in S14, if the center coordinates of all continuous black level regions Rb are not located at the center of all pixel regions, continuous black level regions Rb
Check whether a pair of b exists across the machine width in the center of all pixel areas, and if so, go to S20; if not, go to 3
Proceed to 2B. Here, the center of the total pixel area means the vicinity of the most central pixel of the total pixel area, and the machine width means a relatively small width (here, the number of pixels is 10% or less of the total number of pixels). Naturally, the pixels of this width are at the white level.

At 820, it is checked whether or not these pair of continuous black level regions Rb have a predetermined width (in this case, the number of pixels for each is 20% or less), and if so, a command to turn on is issued.322>
If not, the process advances to S28.

After lighting (322>), it is checked whether all the pixel areas are approximately at the white level (324).Here, the approximate self level includes a continuous black level area Rb of several consecutive pixels or less.If all the pixels If the area is approximately at the white level, a light-off command is issued to the auto light control device to turn off the light, and if not, the process advances to S28.

In S28, the process waits until the end of the above-described even scan period (the start of the next odd scan period), and then returns to 310.

Figure 2 shows the image transport state of the tunnel 5, and Figure 3 shows the bridge girder 6.
shows the adjoining state of .

As shown in Fig. 2, when the tunnel 5 is imaged by the linear image sensor 1b whose pixels are arranged in the horizontal direction,
Does the continuous black level region Rb exist in the central region of the pixel region 1C of the linear image sensor 1b with a width equal to or larger than a predetermined width?
Alternatively, a pair of continuous black level regions Rb are located at the center of all pixel regions and exist with a machine width of a predetermined width or less in between. Here, this aircraft width is the white level area when the tunnel exit is imaged. On the other hand, when passing through a bridge girder that should not be lit (or when passing through a short tunnel), a pair of continuous black level regions Rb exist sandwiching a white level region Rw of a predetermined width or more located at the center of all pixel regions. The tunnel detection device of this embodiment was developed by focusing on the difference in image signals when the tunnel 5 and the bridge girder 6 are imaged by the linear image sensor 1b.

An example of an integrated chip suitable for this embodiment is shown in FIG.

This chip integrates a linear image sensor 11, a microcomputer (or image processing unit) 12 including a buffer 7 memory for one pixel row, and an input/output interface 13. The linear image sensor 11 is preferably of a MOS shift register drive type from the viewpoint of manufacturing process consistency.

The device of this embodiment can also detect raindrops attached to the windshield and the operation of the wiper blade. For example, as shown in FIG. 7, light that passes through raindrops 8 attached to the windshield 7 and forms an image on the linear image sensor 1b forms a pair of white level regions Rb close to the raindrops. This is because the light is refracted and condensed at the left and right inclined portions of the raindrops 8, and the amount of light incident on the linear image sensor 1b increases. On the other hand, the wiper blade 9 forms an image in which a narrow continuous black level region Rb reciprocates over the pixel region tfclc.

Therefore, this device can also detect raindrops and wiper blade operation.

[Second example] A second embodiment of the present invention will be described.

The apparatus of this embodiment is identical to that of FIG. 1, except that
The operation of microcomputer 3 is different.

The operation of the microcomputer 3 of this embodiment is shown in the flowchart of FIG.

That is, when a continuous black level region Rb in a stationary or relatively moving state is detected in S16 or S20, it is determined whether the detected continuous black level region Rb is expanding (that is, whether the distance between the vehicle and the tunnel is decreasing). Find out. To explain specifically, the continuous black level region R extracted this time
The width of b is determined and then compared with the width of the previously extracted continuous black level region Rb (stored in memory). If the width increases by a predetermined width or more, it can be seen that the continuous black level region Rb is being expanded.

Then, if the width of the continuous black level region Rb increases,
If it is lit, the process goes to S22>, otherwise the process goes to 828.

In this way, it is possible to prevent the light from turning on when the vehicle is stopped just before a tunnel or due to a black vehicle in front.

In each of the above embodiments, the threshold voltage Vt of the comparator 2 is kept constant, but the threshold voltage Vref may be changed in conjunction with external illuminance.

[Effects of the Invention] As explained above, the tunnel detection device of the present invention has the advantage that the continuous black level area of a binarized image pickup device of a one-dimensional image device in which pixel rows are arranged horizontally is If it exists with a predetermined width in the center of the imaging area, and if it exists in pairs with a predetermined width on both sides of the predetermined narrow width in the center of the i image area of the -dimensional 1i imaging device, it is determined that it is a tunnel. It has tunnel discrimination means.

Therefore, according to the present invention, high-speed discrimination is possible with a much simpler device configuration than a method of detecting tunnels using a two-dimensional imaging device, and high discrimination accuracy can be obtained.

That is, when the tunnel is sliced into small widths in the horizontal direction, the present invention creates a continuous black level area with a predetermined width in the center of the image transport area, or with a predetermined narrow width in the center of the image transport area (at the exit of the tunnel). We focused on the point that if we identify a tunnel based on these two characteristics, we can identify it quickly and accurately with a simple device configuration. This is what I did.

Furthermore, according to the present invention, image processing of the road surface below the tunnel, various wall surfaces (or the sky) above the tunnel, etc. is not performed, so misclassification due to these surfaces having various image bumps can be prevented. can.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing the first embodiment of the present invention, Fig. 2 is an explanatory drawing showing the tunnel @ image state, Fig. 3 is an explanatory drawing showing the bridge girder transport state, and Fig. 4 is the first embodiment. FIG. 5 is a schematic side view showing the arrangement of the tunnel detection device, FIG. 6 is a schematic plan view of the integrated chip, FIG. 7 is an explanatory diagram showing the raindrop transport state, and FIG. FIG. 9, which is an explanatory diagram showing the wiper blade @ image state, is a partial flowchart explaining the operation of the device of the second embodiment. 1... -dimensional imager 2... Comparator (binarization means)

Claims (1)

[Claims] (1) a one-dimensional imaging device that is mounted on a vehicle in the direction of vehicle travel and captures images in the lateral direction; a binarization means that binarizes a one-dimensional video signal output from the one-dimensional imaging device; tunnel discrimination means for processing the binary video signal outputted from the value conversion means to discriminate a tunnel; are present in a predetermined width at the center of the imaging region of the one-dimensional imaging device, and when they are present in pairs with a predetermined width on both sides of a predetermined narrow width at the center of the imaging region of the one-dimensional imaging device. A tunnel detection device characterized in that it outputs a tunnel detection signal.