JP6139493B2 - License plate detection device and license plate detection method - Google Patents

Description

  The present invention relates to a license plate detection device and a license plate detection method for detecting a license plate of a vehicle based on a plurality of captured images obtained by repeated imaging at a predetermined imaging cycle.

  In recent years, in the field of traffic monitoring and the like, a technique for reading a license plate of a vehicle has been developed as one technique for extracting a specific vehicle. Such a technology for reading a license plate can be used for a toll collection system on a highway, a traffic flow measurement system, an entry / exit vehicle management system such as a parking lot, and the like.

  Further, as such a license plate reading technique, for example, a technique is disclosed in which a processing target region is divided and image processing is performed using a plurality of video cameras in order to process a target range that exceeds a normal image range. (See Patent Document 1).

JP-A-7-105352

  However, since the technique of Patent Document 1 uses a plurality of video cameras, the video camera installation work increases and the cost also increases as compared with the case of using a single video camera. In addition, when performing traffic monitoring, there is a desire to detect license plates of vehicles traveling in all lanes on a road having a plurality of lanes.

  This invention is made | formed in view of such a situation, and it aims at providing the license plate detection apparatus and license plate detection method which can improve cost effectiveness.

A license plate detection device according to an embodiment of the present invention is a license plate that detects a license plate of a vehicle based on a plurality of captured images obtained by repeatedly imaging a predetermined area with a predetermined imaging cycle by a single imaging device. a detection device, the obtained by a single imaging apparatus, from the respective predetermined number of captured images prior Symbol predetermined region is captured over the imaging period of a predetermined number, a part of the predetermined region of the captured image detecting the license plate of the vehicle but on the basis of the partial image cut out by the image extraction portion and, the image cutout unit for cutting out the predetermined number of partial images areas are different for each of the imaging cycle a picture image captured A detecting unit.

A license plate detection method according to an embodiment of the present invention is a license plate that detects a license plate of a vehicle based on a plurality of captured images obtained by repeatedly imaging a predetermined area with a predetermined imaging cycle by a single imaging device. the detection process, the obtained by a single imaging apparatus, from the respective predetermined number of captured images prior Symbol predetermined region is captured over the imaging period of a predetermined number, a part of the predetermined region of the captured image there the steps of cutting out the image cutting unit a partial image of the imaged images and a to the imaging period predetermined number of areas is different for each, the detection unit a number plate of the vehicle based on the partial image extracted Detecting.

  According to the present invention, it is possible to detect a number of vehicle license plates while suppressing costs, and it is possible to improve cost effectiveness.

It is a block diagram which shows an example of a structure of the license plate detection apparatus of 1st Embodiment. It is a schematic diagram which shows an example of a captured image. It is a schematic diagram which shows an example of the partial image which the license plate detection apparatus of 1st Embodiment cuts out. It is a time chart which shows an example of the extraction timing of the partial image by the license plate detection apparatus of 1st Embodiment. It is a block diagram which shows an example of a structure of the license plate detection apparatus of 2nd Embodiment. It is a schematic diagram which shows an example of the partial image which the license plate detection apparatus of 2nd Embodiment cuts out. It is a time chart which shows an example of the extraction timing of the partial image by the license plate detection apparatus of 2nd Embodiment. It is a flowchart which shows an example of the process sequence of the license plate detection apparatus of this Embodiment.

[Description of Embodiment of Present Invention]
(1) A license plate detection apparatus according to an embodiment of the present invention is a license plate detection apparatus that detects a license plate of a vehicle based on a plurality of captured images obtained by repeatedly capturing images at a predetermined imaging cycle. An image cutout unit that cuts out the predetermined number of partial images that are part of the captured image and have different areas for each imaging cycle from each of the captured images obtained over a predetermined number of imaging cycles; A detection unit that detects a license plate of the vehicle based on each partial image cut out by the cutout unit.

(6) A license plate detection method according to an embodiment of the present invention is a license plate detection method for detecting a license plate of a vehicle based on a plurality of captured images obtained by repeatedly capturing images at a predetermined imaging cycle. An image cutout unit that cuts out the predetermined number of partial images that are part of the captured image and have different regions for each imaging cycle, from each of the captured images obtained over a predetermined number of imaging cycles; And a step of detecting a license plate of the vehicle based on each cut out partial image.

  The image cutout unit cuts out a predetermined number of partial images that are part of the captured image and have different areas for each imaging cycle, from each of the captured images obtained over a predetermined number of imaging cycles. For example, when the road is a two-lane road, the captured image is an image of a two-lane road. Further, the predetermined number can be the number of lanes, and in the case of two lanes on one side, the predetermined number is 2. For convenience, the lanes are lane 1 and lane 2. When the imaging cycle is expressed as T1, T2, T3, T4,... In order, the image cutout unit cuts out the partial image including the lane 1 in the imaging cycle T1, and the partial image including the lane 2 in the next imaging cycle T2. cut. Since the predetermined number is 2, in the next imaging cycle T3, the partial image including the lane 1 is cut out after returning to the lane 1, and the partial image including the lane 2 is cut out in the next imaging cycle T4. Thereafter, the same processing is repeated.

  A detection part detects the license plate of a vehicle based on each partial image cut out by the image cutout part. That is, the license plate of the vehicle traveling in the lane 1 is detected based on the partial image cut out in the imaging cycle T1, and the license plate of the vehicle traveling in the lane 2 is detected based on the partial image cut out in the imaging cycle T2. Since the predetermined number is 2, in the next imaging cycle T3, the number plate of the vehicle traveling in the lane 1 is detected based on the cut out partial image, and in the imaging cycle T4, the vehicle travels in the lane 2 based on the cut out partial image. Detect the vehicle license plate. Thereafter, the same processing is repeated.

  With the above-described configuration, it is possible to detect a license plate of a vehicle traveling in all lanes in one direction of the road from a captured image captured by one imaging device (such as a video camera). There is no need to install. Moreover, since a partial image corresponding to each lane is cut out from the captured image, it is only necessary to perform image processing on a partial image smaller than the captured image, as compared with the case where image processing is performed on the entire captured image. The processing load can be reduced, and the apparatus can be prevented from becoming expensive. Thereby, it is possible to detect the number plates of many vehicles while suppressing costs, and to improve cost effectiveness.

(2) The license plate detection apparatus according to an embodiment of the present invention includes an adjustment unit that adjusts an exposure amount at the time of imaging of each of the plurality of captured images for each imaging cycle, and the image cutout unit includes: A partial image whose exposure amount has been adjusted by the adjustment unit is cut out.

  The adjustment unit adjusts the exposure amount at the time of capturing each of the plurality of captured images for each imaging cycle. For example, a lane 1 on a road with two lanes on one side is shaded by a building or other structure, and the image is dark (low brightness), and the other lane 2 is not shaded and is relatively bright. (Brightness is high). In this case, when the partial image including the lane 1 is cut out in the imaging cycle T1, the imaging is performed with the exposure amount E1 in the imaging cycle T1. Further, when a partial image including the lane 2 is cut out in the imaging cycle T2, the imaging is performed with the exposure amount E2 (<E1) in the imaging cycle T2. Thereafter, the same processing is repeated.

  The image cutout unit cuts out the partial image whose exposure amount has been adjusted by the adjustment unit. That is, since the partial image including the lane 1 with low luminance has a large exposure amount at the time of imaging, the luminance of the character portion and the background portion of the license plate is increased by increasing the luminance of the entire image and making the image clear. The difference can be increased and the license plate can be detected with high accuracy. In addition, since the partial image including the lane 2 with high luminance has a small exposure amount at the time of imaging, the contrast can be adjusted to increase the luminance difference between the character portion and the background portion of the license plate. The license plate can be detected well. Thereby, even when the brightness of each lane fluctuates due to outdoor weather changes and changes in sunshine conditions, a partial image whose luminance is adjusted for each lane is cut out, so that the detection accuracy of the license plate can be increased.

(3) In the license plate detection apparatus according to the embodiment of the present invention, the image cutout unit cuts out partial images so as to overlap each partial region cut out in adjacent imaging cycles. is there.

  The image cutout unit cuts out the partial images so as to overlap with partial areas of the partial images cut out at adjacent imaging cycles. For example, when a partial image including the lane 1 is cut out at the imaging cycle T1 and a partial image including the lane 2 is cut out at the imaging cycle T2, each partial image includes a partial region including a lane boundary line between the lane 1 and the lane 2 To overlap. Thus, for example, when the vehicle changes lanes from lane 1 to lane 2, even if the area corresponding to the license plate of the vehicle moves on the lane boundary line, it is included in one of the partial images. And the license plate of the vehicle can be reliably detected.

(4) In the license plate detection device according to the embodiment of the present invention, the image cutout unit cuts out the partial image so that the partial region has a size including an imaging region obtained by imaging the license plate. is there.

  The image cutout section cuts out the partial image so that the partial area has a size including the imaging area obtained by imaging the license plate. Thus, for example, when the vehicle changes lanes from lane 1 to lane 2, the imaging area obtained by imaging the license plate of the vehicle can be included in a partial area of one of the partial images. The license plate can be reliably detected.

(5) A license plate detection device according to an embodiment of the present invention includes a predetermined number of the detection units, and a distribution unit that distributes the predetermined number of partial images cut out by the image cutout unit to the detection units. Is provided.

  A predetermined number of detection units are provided, and the distribution unit distributes the predetermined number of partial images cut out by the image cutout unit to the detection units. For example, if the predetermined number is 2, two detection units are provided. For example, when a partial image including the lane 1 is cut out at the imaging cycle T1 and a partial image including the lane 2 is extracted at the imaging cycle T2, the distribution unit distributes the partial image including the lane 1 to one detection unit, and the lane 2 The partial image to be included is distributed to other detection units. As a result, even when there is a limit on the amount of data sent to one detection unit (for example, the upper limit of the transmission amount), the data can be distributed to a predetermined number of detection units, so that the vehicle number is matched to the imaging cycle. Plate detection processing can be performed.

[Details of the embodiment of the present invention]
(First embodiment)
Hereinafter, the present invention will be described with reference to the drawings illustrating embodiments. FIG. 1 is a block diagram showing an example of the configuration of the license plate detection apparatus 100 according to the first embodiment. An imaging device 200 is connected to the license plate detection device 100. Note that the imaging device 200 may be incorporated into the license plate detection device 100.

  The imaging device 200 is installed at a predetermined position in order to image a vehicle traveling on a road.

  FIG. 2 is a schematic diagram illustrating an example of a captured image. As shown in FIG. 2, when the road is one side two lanes, the picked-up image picked up by the image pickup device 200 can detect the license plates of the two lanes so that the license plates of the vehicles traveling in the lane 1 and the lane 2 can be detected. It is the image which imaged the road.

  The imaging device 200 repeatedly captures images at a predetermined imaging cycle (for example, 30 frames per second). The imaging apparatus 200 outputs a captured image obtained by imaging to the license plate detection apparatus 100 in units of frames. Further, the imaging device 200 adjusts the exposure amount for each imaging cycle in accordance with the exposure amount instruction from the license plate detection device 100 to capture an image. Specifically, the imaging apparatus 200 can adjust the shutter speed for each imaging cycle.

  As shown in FIG. 1, a license plate detection apparatus 100 includes a control unit 10 that controls the entire apparatus, an exposure amount instruction unit 20, a communication unit 30, an image processing unit 40, a plate detection unit 50, a memory 60, and an image processing unit 70. , A plate detector 80, a memory 90, and the like. The image processing unit 40 further includes an image cutout unit 41, a distribution unit 42, and the like. The difference between the image processing unit 40 and the image processing unit 70 is that the image processing unit 40 includes an image cutout unit 41, a distribution unit 42, and the like. In addition, the plate detectors 50 and 80 can have the same configuration.

  The image processing unit 40 acquires the captured image output by the imaging device 200 in units of frames.

  The image cutout unit 41 cuts out a predetermined number of partial images that are part of the captured image and have different areas for each imaging cycle, from each of the captured images obtained over a predetermined number of imaging cycles.

  FIG. 3 is a schematic diagram illustrating an example of a partial image cut out by the license plate detection apparatus 100 according to the first embodiment. The captured image shown in FIG. 3 is the same as FIG. The predetermined number can be the number of lanes, and as shown in FIG. 3, the predetermined number can be two when the road has two lanes on one side. For convenience, the lanes are lane 1 and lane 2. A lane boundary line is provided at the boundary between the lane 1 and the lane 2. As shown in FIG. 3, since the predetermined number is 2, the image cutout unit 41 cuts out the partial image A1 as an image including the lane 1 and cuts out the partial image A2 as an image including the lane 2. The partial images A1 and A2 are both part of the captured image, and both overlap in a partial region including the lane boundary line (a range indicated by reference sign d in FIG. 3).

  FIG. 4 is a time chart illustrating an example of a partial image cut-out timing by the license plate detection apparatus 100 according to the first embodiment. As shown in FIG. 4, the imaging cycle of the captured image is expressed as T1, T2, T3, T4, T5. When the frame rate is 30 frames per second, the time of each cycle is about 33 ms.

  The image cutout unit 41 cuts out the partial image A1 including the lane 1 in the imaging cycle T1, and cuts out the partial image A2 including the lane 2 in the next imaging cycle T2. Since the predetermined number is 2, in the next imaging cycle T3, the partial image A1 including the lane 1 is cut out in the next imaging cycle T3, and the partial image A2 including the lane 2 is extracted in the next imaging cycle T4. Thereafter, the same processing is repeated. In FIG. 4, for convenience, the region of the captured image corresponding to the cutout image is drawn. For example, the partial image A1 is approximately the left half of the captured image, and the partial image A2 is approximately the right half of the captured image.

  The distribution unit 42 distributes the predetermined number of partial images cut out by the image cutout unit 41 to the plate detection unit 80 via the plate detection unit 50 and the image processing unit 70. As described above, when the predetermined number is 2, the license plate detection apparatus 100 includes two (predetermined number) image processing units 40 and 70 and plate detection units 50 and 80. That is, the distribution unit 42 has a function of distributing the partial image A1 including the lane 1 to the one plate detection unit 50 and distributing the partial image A2 including the lane 2 to the other plate detection unit 80.

  More specifically, the image processing unit 40 performs predetermined image processing on the partial image A1 including the lane 1 cut out by the image cutout unit 41 at the imaging cycle T1. Then, the distribution unit 42 outputs the partial image A2 including the lane 2 cut out by the image cutout unit 41 at the imaging cycle T2 to the image processing unit 70. The image processing unit 70 performs predetermined image processing on the partial image A2 including the lane 2 cut out at the imaging cycle T2.

  The image processing unit 40 performs image processing such as noise removal of partial images, extraction of color components and luminance components, brightness (luminance) adjustment, color inversion, gamma correction, gray scale conversion, and required angle rotation processing, for example. The image processing unit 40 performs pre-processing necessary for vehicle detection, such as binarization processing and difference processing for each frame after performing leveling and feature extraction filter processing. The image processing unit 40 performs vehicle detection processing. The image processing unit 40 outputs a detection result to the plate detection unit 50 when a vehicle is detected. Note that the image processing unit 70 performs the same image processing as the image processing unit 40.

  The plate detection unit 50 has a function as a detection unit and detects the position of the license plate of the vehicle, and recognizes characters such as a land code, a classification number, a hiragana, and a serial number based on the detected license plate position. Process. The plate detection unit 50 outputs the license plate information obtained as a result of the character recognition process to the communication unit 30. The plate detection unit 80 performs the same processing as the plate detection unit 50.

  The memory 60 stores predetermined data such as processing results in the image processing unit 40 and the plate detection unit 50. Further, the memory 90 stores predetermined data such as processing results in the image processing unit 70 and the plate detection unit 80.

  The communication unit 30 transmits the license plate information detected by the plate detection units 50 and 80 to an external device. Note that the communication unit 30 can also receive required data from an external device.

  The plate detection units 50 and 80 detect the license plate of the vehicle based on each partial image cut out by the image cutout unit 41. That is, the plate detection unit 50 detects the license plate of the vehicle traveling in the lane 1 based on the partial image A1 cut out in the imaging cycle T1. The plate detector 80 detects the license plate of the vehicle traveling in the lane 2 based on the partial image A2 cut out in the imaging cycle T2. Since the predetermined number is 2, the plate detection unit 50 detects the license plate of the vehicle traveling in the lane 1 based on the cut out partial image A1 in the next imaging cycle T3, and the plate detection unit 80 detects the imaging cycle T4. Then, the license plate of the vehicle traveling in the lane 2 is detected based on the cut out partial image A2. Thereafter, the same processing is repeated. In the example of FIG. 1, a predetermined number of plate detection units are provided. However, the number of plate detection units may be one regardless of the predetermined number.

  With the above-described configuration, it is possible to detect a license plate of a vehicle traveling in all lanes in one direction of the road from a captured image captured by one imaging device 200 (such as a video camera), and thus a plurality of imaging devices. There is no need to install. In addition, since the partial image corresponding to each lane is cut out from the captured image, it is only necessary to perform image processing on a partial image smaller than the captured image, compared to performing image processing on the entire captured image. It is possible to reduce the processing load and to prevent the apparatus from becoming expensive. Thereby, it is possible to detect the number plates of many vehicles while suppressing costs, and to improve cost effectiveness.

  As described above, for example, when the predetermined number is 2, two plate detection units are provided. For example, when the partial image A1 including the lane 1 is cut out at the imaging cycle T1 and the partial image A2 including the lane 2 is cut out at the imaging cycle T2, the plate detection unit 50 displays the license plate for the partial image A1 including the lane 1. A detection process is performed, and the plate detection unit 80 performs a license plate detection process on the partial image A2 including the lane 2. As a result, even when there is a limit on the amount of data sent to one plate detection unit (for example, the upper limit of the data transmission amount between the image processing unit and the plate detection unit), the data is transferred to a predetermined number of plate detection units. Since it can be dispersed, it is possible to detect the license plate of the vehicle in accordance with the imaging cycle.

  The exposure amount instruction unit 20 has a function as an adjustment unit, and adjusts the exposure amount at the time of imaging of each of the plurality of captured images for each imaging period under the control of the control unit 10. For example, in FIG. 3, the lane 1 of a two-lane road is shaded by a building or other structure, the image is dark (low brightness), and the other lane 2 is not shaded. The target image is bright (high brightness). In this case, as shown in FIG. 4, when the partial image A1 including the lane 1 is cut out in the imaging cycle T1, the imaging is performed with the exposure amount E1 in the imaging cycle T1. Further, when the partial image A2 including the lane 2 is cut out in the imaging cycle T2, the imaging is performed with the exposure amount E2 (<E1) in the imaging cycle T2. Thereafter, the same processing is repeated.

  In this case, for example, the control unit 10 determines the lanes of the road according to the date and time, the sunshine condition of each lane of the road for each time zone, the weather conditions, etc. acquired from the external device via the communication unit 30. A command for estimating the brightness and adjusting the exposure amount based on the estimation result may be output to the exposure amount instruction unit 20. In addition, the control unit 10 may estimate the brightness of each lane on the road based on the acquired brightness value of the captured image in a state where the exposure amount of the imaging device 200 is set to the reference value.

  Then, the image cutout unit 41 cuts out the partial images A1 and A2 whose exposure amounts are adjusted. That is, since the partial image A1 including the lane 1 with low luminance has an increased exposure amount at the time of imaging, the luminance of the entire image is increased to make the image clear and the character portion and background portion of the license plate The brightness difference can be increased, and the license plate can be detected with high accuracy. In addition, since the partial image A2 including the lane 2 with high luminance has a reduced exposure amount at the time of imaging, the contrast can be adjusted to increase the luminance difference between the character portion and the background portion of the license plate, The license plate can be detected with high accuracy. Thereby, even when the brightness of each lane fluctuates due to outdoor weather changes and changes in sunshine conditions, a partial image whose luminance is adjusted for each lane is cut out, so that the detection accuracy of the license plate can be increased.

  The image cutout unit 41 cuts out partial images so as to overlap in partial areas of the partial images cut out at adjacent imaging cycles. For example, when the partial image A1 including the lane 1 is cut out at the imaging cycle T1 and the partial image A2 including the lane 2 is cut out at the imaging cycle T2, each of the partial images A1 and A2 is represented by a symbol d in FIG. It overlaps in a partial area including the lane boundary line between the lane 1 and the lane 2. Thus, for example, when the vehicle changes lanes from lane 1 to lane 2, even if the area corresponding to the license plate of the vehicle moves on the lane boundary line, it is included in one of the partial images. And the license plate of the vehicle can be reliably detected.

  In addition, the image cutout unit 41 cuts out the partial image so that the partial area has a size including the imaging area obtained by imaging the license plate. In the example of FIG. 3, the partial image is cut out so that the horizontal length d of the partial region is longer than the size of the license plate (the horizontal length). Thus, for example, when the vehicle changes lanes from lane 1 to lane 2, the imaging area obtained by imaging the license plate of the vehicle can be included in a partial area of one of the partial images. The license plate can be reliably detected.

(Second Embodiment)
In the first embodiment, the predetermined number is 2. That is, the road imaged by the imaging device 200 is a one-lane two-lane road. The number of lanes on the road is not limited to two and may be three or more. In the second embodiment, a case of a three-lane road on one side, that is, a case where the predetermined number is three will be described.

  FIG. 5 is a block diagram showing an example of the configuration of the license plate detection device 110 of the second embodiment. The difference from the license plate detection apparatus 100 of the first embodiment shown in FIG. 1 is that it further includes an image processing unit 71, a plate detection unit 81, and a memory 91, and has a so-called three-system configuration. The image processing unit 71, the plate detection unit 81, and the memory 91 have the same configurations and functions as the image processing unit 70, the plate detection unit 80, and the memory 90, respectively.

  The image cutout unit 41 switches and outputs the cut out partial image to the image processing unit 70 and the image processing unit 71 for each imaging cycle. Details will be described below. Note that the same portions as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted.

  FIG. 6 is a schematic diagram illustrating an example of a partial image cut out by the license plate detection device 110 according to the second embodiment. The predetermined number can be the number of lanes. As shown in FIG. 6, when the road has three lanes on one side, the predetermined number can be three. For convenience, the lanes are lane 1, lane 2, and lane 3. Lane boundaries are provided at the boundary between the lane 1 and the lane 2 and at the boundary between the lane 2 and the lane 3. As shown in FIG. 6, since the predetermined number is 3, the image cutout unit 41 cuts out the partial image B1 as an image including the lane 1, cuts out the partial image B2 as an image including the lane 2, and includes the lane 3. A partial image B3 is cut out as an image. Each of the partial images B1, B2, and B3 is a part of the captured image, and adjacent partial images are overlapped in a partial region including the lane boundary line (a range indicated by reference sign d in FIG. 6). .

  FIG. 7 is a time chart illustrating an example of a partial image cut-out timing by the license plate detection device 110 according to the second embodiment. As shown in FIG. 7, the imaging cycle of the captured image is expressed as T1, T2, T3, T4, T5. When the frame rate is 30 frames per second, the time of each cycle is about 33 ms.

  The image cutout unit 41 cuts out the partial image B1 including the lane 1 in the imaging cycle T1, cuts out the partial image B2 including the lane 2 in the next imaging cycle T2, and includes the lane 3 in the next imaging cycle T3. The image B3 is cut out. Since the predetermined number is 3, in the next imaging cycle T4, the partial image B1 including the lane 1 is cut out in the next imaging cycle T4, and the partial image B2 including the lane 2 is extracted in the next imaging cycle T5. Thereafter, the same processing is repeated. In FIG. 7, for convenience, the region of the captured image corresponding to the extracted image is drawn together. For example, the partial image B1 is approximately one third on the left side of the captured image, the partial image B2 is approximately one third at the center of the captured image, and the partial image B3 is approximately three minutes on the right side of the captured image. It is 1 of.

  The image processing unit 40 performs predetermined image processing on the partial image B1 including the lane 1 cut out by the image cutting unit 41 at the imaging cycle T1. Then, the distribution unit 42 outputs the partial image B2 including the lane 2 cut out by the image cutting unit 41 at the imaging cycle T2 to the image processing unit 70, and the partial image B3 including the lane 3 cut out at the imaging cycle T3. The image is output to the image processing unit 71.

  The image processing unit 70 performs predetermined image processing on the partial image B2 including the lane 2 cut out at the imaging cycle T2. Further, the image processing unit 71 performs predetermined image processing on the partial image B3 including the lane 3 cut out at the imaging cycle T3.

  The plate detection unit 50 performs a license plate detection process on the partial image B1 including the lane 1 cut out at the imaging cycle T1, and the plate detection unit 80 includes the partial image including the lane 2 cut out at the imaging cycle T2. The license plate detection process is performed on B2, and the plate detection unit 81 performs the license plate detection process on the partial image B3 including the lane 3 cut out at the imaging period T3.

  In FIG. 6, the lane 1 of the road with three lanes on one side is behind a building or other structure, the image is dark (the brightness is low), the lane 2 is slightly dark, and the lane 3 is It is assumed that the image is not shaded and the image is relatively bright (high brightness). In this case, as shown in FIG. 7, when the partial image B1 including the lane 1 is cut out in the imaging cycle T1, the imaging is performed with the exposure amount E11 in the imaging cycle T1. Further, when the partial image B2 including the lane 2 is cut out in the imaging cycle T2, the imaging is performed with the exposure amount E12 (<E11) in the imaging cycle T2. Further, when the partial image B3 including the lane 3 is cut out in the imaging cycle T3, imaging is performed with the exposure amount E13 (<E12) in the imaging cycle T3. Thereafter, the same processing is repeated.

  With the above-described configuration, it is possible to detect a license plate of a vehicle traveling in all lanes in one direction of the road from a captured image captured by one imaging device 200 (such as a video camera), and thus a plurality of imaging devices. There is no need to install. In addition, since the partial image corresponding to each lane is cut out from the captured image, it is only necessary to perform image processing on a partial image smaller than the captured image, compared to performing image processing on the entire captured image. It is possible to reduce the processing load and to prevent the apparatus from becoming expensive. Thereby, it is possible to detect the number plates of many vehicles while suppressing costs, and to improve cost effectiveness.

  Further, even when the brightness of each lane fluctuates due to outdoor weather changes and changes in sunshine conditions, a partial image with the brightness adjusted for each lane is cut out, so the license plate detection accuracy can be increased.

  Next, the operation of the license plate detection apparatus of the first embodiment and the second embodiment (hereinafter also referred to as this embodiment) will be described. FIG. 8 is a flowchart showing an example of the processing procedure of the license plate detection apparatus of the present embodiment. In the following description, the subject of processing will be described as the control unit 10 for convenience. The control unit 10 sets the counting code N to 1 (S11), and sets (specifies) the Nth exposure amount (S12). N is an integer from 1 to a predetermined number.

  The control unit 10 acquires a captured image in an arbitrary imaging cycle (S13), and cuts out the Nth partial image (S14). The control unit 10 detects the vehicle based on the cut out partial image (S15), detects the license plate based on the detected vehicle position (S16), and performs character recognition from the detected position of the license plate. (S17), license plate information is detected.

  The control unit 10 determines whether or not N is larger than a predetermined number (S18). If N is not larger than the predetermined number (NO in S18), 1 is added to N (S19), and the processing after step S12 repeat. When N is larger than the predetermined number (YES in S18), the control unit 10 determines whether or not to end the process (S20), and when the process is not ended (NO in S20), the processes after step S11 are repeated. . When the process ends (YES in S20), the control unit 10 ends the process.

  As described above, according to the present embodiment, a single imaging device captures all lanes of a road to detect a vehicle license plate and reduces the processing load of image processing when detecting a license plate. Can be cost-effective. In addition, since the brightness of the partial image for each lane can be individually adjusted, the detection accuracy of the license plate can be improved even when there is a change in weather conditions or sunshine conditions.

  In the above-described embodiment, the frame rate of the imaging apparatus is 30 frames per second. However, the frame rate is not limited to this, and may be, for example, a frame rate of 60 frames per second.

  The disclosed embodiments are to be considered in all respects as illustrative and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

DESCRIPTION OF SYMBOLS 10 Control part 20 Exposure amount instruction | indication part 30 Communication part 40, 70, 71 Image processing part 41 Image extraction part 42 Distribution part 50, 80, 81 Plate detection part 60, 90, 91 Memory 100 Number plate detection apparatus 200 Imaging device

Claims (6)

A license plate detection device that detects a license plate of a vehicle based on a plurality of captured images obtained by repeatedly imaging a predetermined area with a predetermined imaging cycle by one imaging device,
Obtained by the single image pickup device, from said respective predetermined number of captured images prior Symbol predetermined region is captured over the imaging period of a predetermined number, images a part of the predetermined region of the captured image was captured And an image cutout unit that cuts out the predetermined number of partial images having different regions for each imaging cycle,
A license plate detection apparatus comprising: a detection unit that detects a license plate of a vehicle based on each partial image cut out by the image cutout unit. An adjustment unit that adjusts an exposure amount at the time of imaging each of the plurality of captured images for each imaging cycle;
The image cutout part is
The license plate detection apparatus according to claim 1, wherein a partial image whose exposure amount has been adjusted by the adjustment unit is cut out. The image cutout part is
3. The license plate detection apparatus according to claim 1, wherein the partial images are cut out so as to overlap each other in a partial area of each partial image cut out at an adjacent imaging cycle. The image cutout part is
The license plate detection apparatus according to claim 3, wherein the partial image is cut out so that the partial area has a size including an imaging region obtained by imaging the license plate. The detection unit includes the predetermined number,
5. The license plate detection apparatus according to claim 1, further comprising: a distribution unit that distributes the predetermined number of partial images cut out by the image cutout unit to the detection units. A license plate detection method for detecting a license plate of a vehicle based on a plurality of captured images obtained by repeatedly imaging a predetermined area with a predetermined imaging cycle by one imaging device,
Obtained by the single image pickup device, from said respective predetermined number of captured images prior Symbol predetermined region is captured over the imaging period of a predetermined number, images a part of the predetermined region of the captured image was captured An image cutout unit cutting out the predetermined number of partial images having different areas for each imaging cycle;
A detecting unit for detecting a license plate of a vehicle based on each cut-out partial image.

Images