JP2014191763A - Device, system, program, and method for recognizing vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress deterioration in vehicle recognition performance by erroneously detecting a vehicle passing on another lane adjacent to a lane to be imaged as a vehicle passing on the lane to be imaged.SOLUTION: An inventive recognition device includes a specification part, a judgement part, and an exclusion part. The specification part specifies a symbol of a number plate of a vehicle from an image imaged by a second imaging apparatus which is intended to image a second lane. The second imaging apparatus is installed at a position which is more distant to the second lane adjacent to a first lane than to the first lane which is an imaging target of a first imaging apparatus. The judgement part determines whether a large vehicle is detected from an image imaged by the first imaging apparatus during a predetermined period based on a time at which an image of the vehicle is imaged when failing in specification of the symbol. The exclusion part excludes the image of the vehicle imaged by the second imaging apparatus from a case example indicating failing in specification of the symbol of the number plate when the large vehicle is detected.

Description

  The present invention relates to a vehicle recognition device, system, program, and method.

  Currently, an imaging device such as a camera is installed on a road, and a vehicle passing through the road is detected by a recognition device based on an image captured by the imaging device. Moreover, the recognition apparatus recognizes the symbol described in the number plate of a vehicle from the image in which the vehicle was detected. The vehicle information obtained from these images is used for, for example, vehicle entry / exit management in parking lots, destination instructions for each vehicle in factories, etc., search for arranged vehicles in organizations such as the police, and the like.

  In this regard, a technique is known in which an allowable value of the reading rate is determined based on a function of the number of passing vehicles, and the allowable value is sequentially changed according to the number of passing vehicles. In addition, a technique for reading a license plate of a vehicle is known. (For example, see Patent Document 1 and Patent Document 2)

JP 2004-258853 A Special table 2004-525447 gazette

  However, a vehicle passing another adjacent lane may be reflected in an image obtained by capturing the lane to be imaged by the imaging device, and may be erroneously detected as a vehicle passing through the lane to be imaged. Such erroneous detection is not preferable because it reduces the recognition performance of the vehicle by the recognition device.

  In one aspect, an object of the present invention is to suppress a decrease in recognition performance of a vehicle by erroneously detecting a vehicle passing through another lane adjacent to the imaging target lane as a vehicle passing through the imaging target lane. To do.

  The recognition device according to one aspect of the present invention includes a specifying unit, a determination unit, and an exclusion unit. The specifying unit specifies a symbol described on the license plate of the vehicle from the image of the vehicle captured by the second imaging device that captures the second lane. Here, the second imaging device is installed at a position where the distance to the second lane adjacent to the first lane is farther than the distance to the first lane that is the imaging target of the first imaging device. Has been. The determination unit satisfies a predetermined condition from the image captured by the first imaging device during a predetermined period specified based on the time when the image of the vehicle is captured when the identification unit fails to identify the symbol. It is determined whether or not a vehicle has been detected. When the determination unit determines that the vehicle is detected, the exclusion unit excludes the vehicle image captured by the second imaging device from an example indicating that the identification of the license plate symbol has failed.

  It is possible to suppress a reduction in the recognition performance of the vehicle by erroneously detecting a vehicle passing through another lane adjacent to the imaging target lane as a vehicle passing through the imaging target lane.

1 is a diagram illustrating a system for recognizing a vehicle according to some embodiments. FIG. It is a figure which shows the example which misrecognizes the vehicle which passes an adjacent lane as a vehicle which passes the lane of recognition object. It is a figure which shows the example in which the upper part of the large vehicle which drive | works a 1st lane was reflected in the image of the 2nd imaging device which makes a 2nd lane imaging object. It is a figure which illustrates the functional block structure of the recognition apparatus which concerns on some embodiment. It is a figure which illustrates the recognition information which concerns on 1st Embodiment. It is an operation | movement flowchart which illustrates the vehicle recognition process of the 1st lane which concerns on 1st Embodiment. It is an operation | movement flowchart which illustrates the vehicle recognition process of the 2nd lane which concerns on 1st Embodiment. It is an operation | movement flowchart which illustrates the notification process of the recognition information which concerns on 2nd Embodiment. It is a figure which illustrates the functional block structure of the center apparatus which concerns on some embodiment. It is an operation | movement flowchart which illustrates the exclusion process by the center apparatus which concerns on 3rd Embodiment. It is a figure which illustrates the hardware constitutions of the computer for implement | achieving the recognition apparatus which concerns on some embodiment. It is a figure which illustrates the hardware constitutions of the computer for implement | achieving the center apparatus which concerns on some embodiment.

Hereinafter, some embodiments of the present invention will be described in detail with reference to the drawings. In addition, the same code | symbol was attached | subjected to the corresponding element in several drawing.
FIG. 1 is a diagram illustrating a system 1 for recognizing a vehicle according to some embodiments. In the example illustrated in FIG. 1, the system 1 includes a center device 2, a recognition device 3, and an imaging device 4. The imaging device 4 is provided for each lane to be imaged, for example, and images the lane to be imaged and acquires an image of the vehicle passing there. The recognizing device 3 is connected to the imaging device 4, captures an image captured by the imaging device 4, and detects a vehicle from the captured image. Moreover, the recognition apparatus 3 performs the process which acquires the information of the vehicle reflected on the image, when a vehicle is detected in the image. In this recognition processing, for example, information such as a symbol written on the license plate and the color of the vehicle body is recognized. The recognized vehicle information is notified to the center apparatus 2 via the network 5.

  When there are a plurality of lanes to be imaged at the installation positions of the recognition device 3 and the imaging device 4, for example, as shown in the installation position 1 in FIG. The information from the imaging device 4 may be processed by one recognition device 3 and transmitted to the center device 2. In another example, as illustrated by the installation position 2 in FIG. 1, a set of the recognition device 3 and the imaging device 4 may be installed for each lane to be imaged. In this case, the plurality of recognition devices 3 installed at one installation position may be connected to each other.

  The center device 2 collects vehicle information transmitted from the recognition device 3, and the information includes, for example, vehicle entry / exit management in parking lots, destination instructions for each vehicle in factories, and police organizations. This is useful for searching for arranged vehicles.

In addition, in order to evaluate the recognition performance of the vehicle by the system 1 for recognizing the vehicle as described above, the symbol recognition success rate described in the license plate is used as an example of the recognition performance index. The recognition success rate of the license plate symbol is calculated by the following equation, for example.
(Number of successful recognition of license plate symbols) = (Number of successful recognition of license plate symbols among vehicles recognized by recognition device) / (Number of vehicles recognized by recognition device)

  Here, (the number of vehicles recognized by the recognition device) means, for example, that the recognition device 3 connected to the imaging device 4 performs a vehicle recognition process on the image captured by the imaging device 4. As a result, the number of vehicles recognized within a predetermined period. Further, (the number of successfully recognized license plate symbols among the vehicles recognized by the recognition device) is, for example, further described in a license plate (vehicle number mark) with respect to the image in which the vehicle is recognized. This is the number of symbols that have been successfully recognized. The recognition device 3 performs license plate symbol recognition on an image in which the vehicle has been recognized, but the recognition result may not be able to obtain reliability higher than a predetermined value. Examples of cases where such reliability is not obtained include, for example, if the number plate is attached with dirt, etc., and the symbol to be recognized cannot be read, or if the number plate has a cover. Is mentioned. Then, the recognition device 3 is configured to determine that the recognition result is unknown (symbol recognition fails) when reliability higher than a predetermined level cannot be obtained in the recognition result.

  Therefore, the above (the number of successful symbol recognition results among the vehicles recognized by the recognition device) is, in other words, the recognition device 3 did not make the recognition result unknown during symbol recognition (for symbol recognition). Number that did not fail). Note that the above-described successful recognition means that the recognition device 3 has recognized the symbol with a certain level of reliability, and the recognized symbol matches the actual symbol of the imaged vehicle. It doesn't mean to do. That is, the resulting recognized symbol may not match the actual symbol of the imaged vehicle.

  The (number plate symbol recognition success rate) calculated as described above is used as an index for evaluating the symbol recognition performance of the system 1 by the system 1 distributor. Alternatively, it is used as an index for evaluating the recognition performance of the vehicle in the lane where the user of the system 1 wants to recognize the vehicle, for example.

  Here, for example, in the recognition of the vehicle by the system 1 described above, a vehicle passing through the adjacent lane may be erroneously recognized as a vehicle passing through the recognition target lane. FIG. 2 shows an example in which when an imaging device installed on a road captures a vehicle passing through an imaging target lane, a vehicle passing through an adjacent lane is erroneously recognized as a vehicle passing through a recognition target lane. FIG. FIG. 2 shows two lanes 200 (first lane 201 and second lane 202) through which the vehicle passes. In addition, two imaging devices 4 (for example, the first imaging device 211 and the second imaging device 212) are installed on a pillar 203 such as a signal pole along the sidewalk. For example, the first imaging device 211 has the first lane 201 as an imaging target, and is installed at an angle so that a vehicle traveling in the first lane 201 can be imaged in the vicinity of the imaging area 205 (1). In addition, the second imaging device 212 is set at an angle so that, for example, the second lane 202 is an imaging target, and a vehicle traveling in the second lane 202 can be imaged in the vicinity of the imaging area 205 (2). For example, as described above, it is often performed that a plurality of imaging devices 4 are installed at a predetermined installation position such as the pillar 203 and a plurality of lanes are imaged from the predetermined position.

  However, when imaging the lane to be imaged, the imaging device 4 may be installed at a position where the image must be captured across another lane on the near side of the lane to be imaged. In this case, a vehicle passing through another lane on the near side may be reflected in the imaging range of the imaging device 4. For example, the second imaging device 212 targets the second lane 202, but in order to capture the second lane 202 from the installation position of the second imaging device 212, it is necessary to capture across the first lane 201. There is. For this reason, the imaging range of the second imaging device 212 includes a space immediately above the first lane 201. For example, it is assumed that the large vehicle 250 passes through the first lane 201 on the near side from the installation position of the second imaging device 212 closer to the second lane 202 to be imaged. In such a case, for example, the upper portion 206 of the large vehicle 250 may be reflected in the image captured by the second imaging device 212.

  FIG. 3 is a diagram illustrating an example in which the upper part 206 of the large vehicle 250 traveling in the first lane 201 is reflected in the image 300 of the second imaging device 212 that captures the second lane 202. In the image 300, the hatched area represents the second lane 202. In the image 300, an upper portion 206 of the large vehicle 250 traveling in the first lane 201 is reflected. Since the large vehicle 250 is moving on the first lane 201, the large vehicle 250 also moves within the imaging range 310 by the second imaging device 212. Therefore, for example, the recognition device 3 that detects the vehicle based on the image captured by the second imaging device 212 erroneously recognizes the large vehicle 250 that passes through the first lane 201 as a vehicle that passes through the second lane 202. . That is, it is not preferable because a vehicle traveling in a lane other than the recognition target lane is erroneously recognized as a vehicle traveling in the recognition target lane.

  In addition, as described above, during vehicle recognition, a license plate may be recognized from a captured image of the vehicle to identify a symbol. As shown in FIG. 3, a large image is shown in the image. It is the upper part 206 of the vehicle 250, and the license plate is not shown. For this reason, the process of recognizing the license plate and specifying the symbol also fails. As a result, the fact that a vehicle traveling in a vehicle other than the recognition target lane is erroneously recognized as a vehicle traveling in the recognition target lane is the above-described (number plate symbol recognition success rate). ). However, since this decrease in the (number plate symbol recognition success rate) is caused regardless of the vehicle passing through the lane to be recognized, it is noise that reduces the accuracy of the (number plate symbol recognition success rate). . Therefore, it is desirable to suppress erroneous recognition of a vehicle traveling in the lane 200 that is not the imaging target as a vehicle traveling in the lane 200 that is the imaging target.

  A misrecognition that recognizes such a vehicle traveling in the adjacent lane as a vehicle traveling in the imaging target lane is, for example, the installation position of the imaging device 4 from the front of the vehicle traveling in the imaging target lane. This can be dealt with by arranging it at a position where it can be imaged at an angle. However, since the vehicle recognition system 1 may be used, for example, for searching for arranged vehicles in an organization such as the police, it may be desirable to install the imaging device 4 at a position that is difficult to find from the driver of the vehicle. is there. Moreover, in order to install the imaging device 4 at a position where a vehicle traveling in the lane to be imaged can be imaged from the front, additional equipment may be required. For this reason, in view of cost, it is preferable to install on a pillar 203 such as an existing signal pole along the road. For example, the installation position of the imaging device 4 may be restricted due to such circumstances, and there are situations where it is difficult to select an imaging angle in such a case.

  Thus, in the system 1 according to some embodiments, when there is a lane that is adjacent to the side closer to the distance closer to the imaging target lane from the installation position of the imaging device, the recognition device 3 is the imaging target. Suppose we failed to recognize the lane vehicle symbol. In this case, in a predetermined period based on the time when the image of the vehicle that failed to recognize the symbol was captured, a large vehicle is captured in the captured image of the imaging device that targets the lane adjacent to the nearer side in front. Determine if it is visible. And if a large vehicle was shown, the failure in recognition is likely due to the fact that a large vehicle passing through the lane on the near side was reflected. Are excluded from cases that indicate that the identification of the symbol has failed. For this reason, it is possible to suppress erroneous detection of a vehicle traveling in another lane 200 in front of which is not an imaging target as a vehicle traveling in the imaging target lane 200, and due to the erroneous detection (of the number plate symbol) A reduction in recognition success rate can be prevented.

  In the following description of the embodiment, when two vehicles are recognized in two adjacent lanes using two imaging devices installed at predetermined installation positions, the installation is performed in two adjacent lanes. A lane that is closer to the distance from the position is defined as a first lane 201. An imaging device that targets the first lane 201 on the side closer to the distance from the installation position is referred to as a first imaging device 211. On the other hand, of the two adjacent lanes, the lane on the far side from the installation position is defined as the second lane 202. An imaging device that targets the second lane 202 on the far side from the installation position is referred to as a second imaging device 212. The two adjacent lanes may be, for example, two lanes on one side, two adjacent lanes on three or more lanes, or opposite lanes on one lane on one side. . The predetermined installation position may be, for example, a traffic light or a utility pole.

  FIG. 4 is a diagram illustrating a functional block configuration of the recognition device 3 according to some embodiments including first and second embodiments described later. The recognition device 3 includes a control unit 400 and a storage unit 410, for example. The control unit 400 includes functional units 401 such as a specifying unit 411, a determining unit 412, and an excluding unit 413, for example. The storage unit 410 of the recognition device 3 stores, for example, information such as a program 420 and recognition information 500 described later. The control unit 400 of the recognition device 3 functions as the functional unit 401 such as the specifying unit 411, the determination unit 412, and the exclusion unit 413 by reading and executing the program 420. Details of each functional unit 401 and information stored in the storage unit 410 will be described later.

  FIG. 5 is a diagram illustrating recognition information 500 according to the first embodiment. The recognition information 500 is information that is generated when the recognition device 3 detects a vehicle with respect to an image captured by the imaging device 4, for example, and is recorded in the storage unit 410 when the vehicle is detected. In the recognition information 500, for example, information regarding the detected vehicle is registered. The recognition information 500 includes, for example, an item 501 and a value 502 for the item. In the example of FIG. 5, the recognition information 500 includes, as an item 501, an identification number, a passage date and time, a point, a lane, a land transportation branch code, a classification number, a usage code, and a series designation number, a plate size, a plate type, a vehicle still image, and a vehicle type. , And body color. Item 501: The identification number is assigned to the detected image of the vehicle when the recognition device 3 executes processing for detecting the vehicle on the image captured by the imaging device 4 and the vehicle is detected. This is a number for identifying the recognition information 500 to be recognized. FIG. 5 shows an example in which a serial number of 1 to 10 is assigned as an identification number to an image in which a vehicle is detected, and the value is registered as a value 502. In item 501: passage date and time value 502, the date and time when the imaging device 4 captured the image in which the vehicle was detected, which triggered the generation of the recognition information 500, is registered. The item 501: point is information about the installation position of the imaging device 4 connected to the recognition device 3, and the name 502 such as an address indicating the installation position is registered in the value 502. In another embodiment, information such as latitude and longitude may be registered in the item 501: point. In item 501: lane value 502, information for identifying the lane to be imaged by the imaging device 4 connected to the recognition device 3 is registered.

  The land transportation branch office code, classification number, usage code, and series designation number of the item 501 are symbols written on a Japanese license plate. The land transportation branch office code is, for example, a symbol indicating a place name of a transportation branch office that issued the license plate or an automobile inspection registration office. The classification number is, for example, a symbol indicating classification according to the type and use of the automobile. The application code is, for example, a symbol indicating classification according to the purpose of use of the automobile. The series designation number is, for example, a symbol represented by an Arabic numeral of four digits or less. If the symbol written on the license plate is, for example, “Shiga 500 No. 1”, the land transportation branch code value 502 may be “Shiga”. Further, the classification number value 502 may be “500”. The usage code value 502 may be “no”. The value 502 of the serial designation number may be “••• 1”. For example, the symbol information described on the license plate as described above is registered in the recognition information 500. Note that the symbols on the license plate may have different formats depending on, for example, the country, so depending on the format, the Land Transport Bureau code, classification number, usage code, and series designation number Other information may be registered in the recognition information 500 instead.

  The plate size of the item 501 is information indicating the size of the license plate. For example, in Japan, there are license plates of different sizes such as a large number mark (also referred to as a large plate), a medium size number mark (also referred to as a medium plate), and a small number mark (also referred to as a small plate). The large plate is, for example, the size of a license plate attached to a large vehicle that satisfies a predetermined condition. The middle plate is, for example, the size of a license plate attached to a normal passenger car. The small plate is, for example, the size of a license plate attached to a motorcycle or the like. The large plate is larger than the middle plate and the small plate. Further, the size of the middle plate is larger than that of the small plate. The plate size value 502 of the item 501 includes, for example, the license plate size recognized from the image detected by the vehicle among the three sizes of the large plate, the middle plate, and the small plate. If it corresponds to any of the sizes, information on the corresponding size is registered. In the plate type value 502 of the item 501, for example, the type of a license plate for private use or business use is registered. In Japan, license plate usage codes are allocated according to usage for personal use, etc. Therefore, information for personal use or business use may be registered in the value 502 of the plate type based on the recognition result of the usage code. In the vehicle still image value 502 of the item 501, a storage location (for example, a path and a file name) of an image in which the vehicle is detected, which has triggered the generation of the recognition information 500, is registered. In the vehicle type value 502 of the item 501, information on whether the vehicle detected from the image of the vehicle still image of the item 501 is, for example, large or medium is registered. An example of how to determine whether the vehicle detected from the image is large or medium will be described later. In the body color value 502, the color of the vehicle body detected from the vehicle still image of the item 501 is registered. For example, the recognition information 500 including the above information may be generated when a vehicle is detected from an image captured by the imaging device 4. Note that the information included in the recognition information 500 is not limited to the above, and some of the information described above may not be included, or may include additional additional information.

  FIG. 6 is an operation flowchart illustrating the vehicle recognition process of the first lane 201 executed by the control unit 400 of the recognition device 3 according to the first embodiment. The operation flow of the vehicle recognition process for the first lane 201 in FIG. 6 is implemented, for example, when the control unit 400 of the recognition device 3 reads and executes the program 420 stored in the storage unit 410. In one embodiment, the recognition apparatus 3 starts the vehicle recognition process in FIG. 6 when an instruction to start the vehicle recognition process is input.

  In step S <b> 601, the control unit 400 of the recognition device 3 captures an image captured through the first imaging device 211 that captures the first lane 201. The first lane 201 is, as described above, of two adjacent lanes when using two imaging devices installed at a predetermined installation position to recognize a vehicle in two adjacent lanes. The lane is on the side closer to the distance from the installation position. The first imaging device 211 is an imaging device that targets the first lane 201 on the side closer to the distance from the installation position.

  Subsequently, in step S602, the control unit 400 of the recognition device 3 detects the vehicle from the captured image. The detection of the vehicle may be performed, for example, by detecting a moving object by taking a difference between successive images. In step S603, the control unit 400 of the recognition device 3 determines whether a vehicle is detected. If no vehicle is detected (NO in step S603), the flow returns to step S601. On the other hand, if a vehicle is detected (YES in step S603), the flow proceeds to step S604.

  In step S604, the control unit 400 of the recognition device 3 generates recognition information 500 for registering the detected vehicle information, and the generated recognition information 500 includes a license plate for the image in which the vehicle is detected. Register the information obtained by recognizing symbols. The license plate symbol recognition may be executed by, for example, extracting a license plate region of a vehicle from a captured image and performing character recognition on the image region. As described above, the symbols written on the license plate include, for example, a land transportation branch office code, a classification number, a use code, and a series designation number in Japan. The control unit 400 of the recognition apparatus 3 acquires the information by recognizing the license plate symbol and registers it in the recognition information 500. Note that the information described in the license plate has a different format depending on, for example, the country, and therefore different information may be registered in the recognition information 500 depending on the format. For example, when symbol recognition fails in step S604, information indicating that symbol recognition has failed may be registered in the recognition information 500. In this case, information indicating that the symbol recognition has failed may be registered in the recognition information 500, for example, in the land transportation branch office code, the classification number, the usage code, and the series designation number.

  In step S605, the control unit 400 of the recognition device 3 acquires information about the vehicle shown in the image from which the vehicle is detected. In this process, the control unit 400 of the recognition device 3 specifies, for example, the vehicle type (for example, large size, medium size, etc.) of the vehicle and registers it in the item 501 of the recognition information 500: the value 502 of the vehicle type. Further, the control unit 400 of the recognition apparatus 3 may further acquire other information included in the item 501 of the recognition information 500 and register it in the value 502 of the recognition information 500.

  The vehicle type may be specified as follows, for example. For example, in Japan, license plates with different sizes, colors, symbols, and the like are used according to the weight, loading capacity, and capacity of vehicles. For example, a large vehicle that satisfies a predetermined condition is provided with a number plate having a size called a large number mark (also referred to as a large plate). This large number mark has a size called a medium number number mark (also referred to as a middle plate) attached to a normal passenger car or a size larger than a small number mark (also referred to as a small plate) attached to a small car or a motorcycle. Yes. And the size of the number plate of a vehicle can be specified from the image from which the vehicle was detected. For this reason, the control unit 400 of the recognition device 3 may register the large model in the item 501 of the recognition information 500 and the value 502 of the model if the size of the license plate specified from the image is large. . Further, the control unit 400 of the recognition device 3 may register the vehicle type as a medium model if the size of the license plate specified from the image is a medium plate, and register the medium model in the item 501 of the recognition information 500: the value 502 of the vehicle type. .

  In another embodiment, the control unit 400 of the recognition device 3 extracts an area in which the vehicle is captured from an image in which the vehicle is detected. If the width of the area is equal to or larger than a predetermined size, the recognition information The item 501 of 500 may be registered as large in the value 502 of the vehicle type. Alternatively, the control unit 400 of the recognition device 3 determines that the extracted vehicle area is a predetermined area or more, the height of the vehicle area is a predetermined height or more, or the depth of the vehicle area is a predetermined length. If it is equal to or larger than this, it may be registered as large in the item 501 of the recognition information 500 and the value 502 of the vehicle type. Furthermore, by combining these determinations, when a large vehicle is identified by a plurality of determinations, the large-scale vehicle may be registered in the item 501 of the recognition information 500: the value 502 of the vehicle type. Similarly, if the license plate size is the middle plate, the vehicle type may be specified as the middle size, or the width, area, height, or depth of the vehicle area extracted from the image is within a predetermined range. If it is within, it may be configured to be identified as a medium size. Furthermore, these determinations may be combined, and when a medium-sized vehicle is identified by a plurality of determinations, the medium 501 may be registered in the item 501 of the recognition information 500: the value 502 of the vehicle type.

  In step S <b> 605, the control unit 400 of the recognition device 3 may acquire other information included in the item 501 of the recognition information 500 in addition to the vehicle type, and register it in the value 502 of the recognition information 500. For example, in the example of FIG. 5, the recognition information 500 further includes information on an identification number, a passage date and time, a point, a lane, a plate size, a plate type, a vehicle still image, and a vehicle body color. 400 acquires these pieces of information and registers them in the recognition information 500.

  For example, the control unit 400 of the recognition device 3 registers, for example, a serial number incremented by 1 every time a vehicle is detected in the item 501: identification number. Moreover, the control part 400 of the recognition apparatus 3 registers the imaging time when the image in which the vehicle was detected was imaged, for example in the item 501: passage date. The control unit 400 of the recognition device 3 registers the name of the location of the installation position of the connected imaging device 4 in the item 501: location, and the imaging target of the connected imaging device 4 in the item 501: lane. Register information that identifies the lane of the vehicle. In addition, the name of the installation point of the connected imaging device 4 and information for specifying the lane may be stored in the storage unit 410 in advance, for example. Further, the control unit 400 of the recognition apparatus 3 specifies and registers the number plate size in the item 501: plate size from the image in which the vehicle is detected in step S603. In the example of FIG. 5, when the size of the license plate is a medium plate or a large plate, medium size or large size information corresponding thereto is registered. Furthermore, the control unit 400 of the recognition apparatus 3 registers personal or business information in the item 501: plate type based on, for example, the item 501: application code. The control unit 400 of the recognition apparatus 3 registers, for example, the storage location (for example, the path and file name) of the image in which the vehicle is detected in step S603 in the item 501: vehicle still image, and the vehicle that appears in the image. Is registered in item 501: body color.

  For example, as described above, the control unit 400 of the recognition device 3 registers the vehicle type and other information in the recognition information 500 in step S605. In step S606, the control unit 400 of the recognition device 3 transmits the recognition information 500 in which information about the detected vehicle is registered to the center device 2, and the flow returns to step S601.

  For example, information about vehicles passing through the lane to be imaged is collected by the center device 2 through the processing described above.

  Next, with reference to FIG. 7, a vehicle recognition process for the second lane 202 executed by the control unit 400 of the recognition apparatus 3 according to the first embodiment will be described. FIG. 7 is an operation flowchart illustrating the vehicle recognition process of the second lane 202 according to the first embodiment. The operation flow of the vehicle recognition process for the second lane 202 in FIG. 7 is implemented, for example, when the control unit 400 of the recognition device 3 reads and executes the program 420 stored in the storage unit 410. In one embodiment, the recognition apparatus 3 starts the vehicle recognition process in FIG. 7 when an instruction to start the vehicle recognition process is input. In the operation flow in FIG. 7, the processes in steps S701 to S705 may correspond to the processes in steps S601 to S605 in FIG. For example, a process similar to the process executed in steps S701 to 705 with the second lane 202 as an imaging target and the first lane 201 as an imaging target in steps S601 to 605 may be executed. The second lane 202 is, as described above, of two adjacent lanes when the two imaging lanes are used to recognize a vehicle in two adjacent lanes. The lane is on the far side from the installation position. The second imaging device 212 is an imaging device that targets the second lane 202 on the far side from the installation position.

  In step S <b> 701, the control unit 400 of the recognition device 3 captures an image captured through the second imaging device 212 that captures the second lane 202. Subsequently, in step S702, the control unit 400 of the recognition device 3 detects the vehicle from the captured image. The detection of the vehicle may be performed, for example, by detecting a moving object by taking a difference between successive images. In step S703, the control unit 400 of the recognition device 3 determines whether a vehicle is detected. If no vehicle is detected (NO in step S703), the flow returns to step S701. On the other hand, if a vehicle is detected (YES in step S703), the flow proceeds to step S704.

  In step S704, the control unit 400 of the recognition device 3 generates recognition information 500 for registering the detected vehicle information, and the generated recognition information 500 includes a license plate for the image in which the vehicle is detected. Register the information obtained by recognizing symbols. In addition, when the recognition of the license plate symbol has failed, information indicating that the symbol recognition has failed may be registered in the recognition information 500. In this case, information indicating that the symbol recognition has failed may be registered in the recognition information 500, for example, in the land transportation branch office code, the classification number, the usage code, and the series designation number.

  In step S705, the control unit 400 of the recognition device 3 acquires information about the vehicle shown in the image from which the vehicle is detected. In this process, the control unit 400 of the recognition device 3 specifies, for example, the vehicle type (for example, large size, medium size, etc.) of the vehicle and registers it in the item 501 of the recognition information 500: the value 502 of the vehicle type. Further, the control unit 400 of the recognition apparatus 3 may further acquire other information included in the item 501 of the recognition information 500 and register it in the value 502 of the recognition information 500.

  In step S706, the control unit 400 of the recognition apparatus 3 determines whether or not the recognition of the symbol described on the license plate in step S704 has been successful. If the recognition of the symbol described on the license plate is successful (YES in step S706), the flow proceeds to step S709. On the other hand, when the control unit 400 of the recognition device 3 fails to recognize the symbol described on the license plate (NO in step S706), the flow proceeds to step S707. As described above, the successful recognition of the symbol written on the license plate means that the recognition device 3 can recognize the license plate symbol with a certain level of reliability without making the recognition result unknown. It represents that. Also, the resulting recognized symbol may not match the actual symbol of the imaged vehicle.

  In step S707, the control unit 400 of the recognition device 3 acquires the recognition information 500 of the adjacent first lane 201 in a predetermined period specified based on the time when the image in which the vehicle is detected in step S703 is captured. . For example, the control unit 400 of the recognition device 3 reads the recognition information 500 of the first lane 201 stored in the storage unit 410 to acquire the recognition information 500 of the adjacent first lane 201 in a predetermined period. Also good. For example, when the vehicle detected in step S703 is a large vehicle passing through the adjacent first lane 201, the large vehicle is in the first lane 201 and the first imaging device 211 is in this predetermined period. Is set to a period during which the image is captured and detected. For example, in the example of FIG. 2, the large vehicle 250 is shown at the illustrated position, and the upper portion 206 of the large vehicle 250 appears in the image of the second lane 202. The image is picked up and detected by one image pickup device 211. Then, by taking these detected positions, the vehicle speed, etc. into consideration, a period during which the large vehicle 250 reflected in the captured image of the second imaging device 212 will be imaged and detected by the first imaging device 211 Can be estimated. For example, in one embodiment, this estimated period may be set as a predetermined period. The predetermined period is, for example, a period of about 1.6 to 0.4 seconds, which takes about 0.8 to 0.2 seconds before and after the time when the image where the vehicle is detected in step S703 is captured. There may be. In another example, considering the moving time of the large vehicle 250 from the position of the imaging area 205 (1) to the illustrated position where the upper portion 206 is reflected, the central time of the predetermined period is set in step S703. It may be set to be shifted from the time when the image where the vehicle is detected is captured.

  If the recognition information 500 of the adjacent first lane 201 is acquired in step S707, the control unit 400 of the recognition device 3 determines that the vehicle type is large in the acquired recognition information 500 of the adjacent first lane 201 in step S708. It is determined whether or not the recognition information 500 is included. When the recognition information 500 determined that the vehicle type is large is not included (NO in step S708), the flow proceeds to step S709. In step S709, the control unit 400 of the recognition apparatus 3 transmits recognition information 500 including information about the vehicle registered in steps S704 to 705 to the center apparatus 2, and the flow returns to step S701.

  On the other hand, when the recognition information 500 determined that the vehicle type is large is included (YES in step S708), the flow proceeds to step S710. In step S710, the control unit 400 of the recognition apparatus 3 discards the recognition information 500 including the image in which the vehicle is detected in step S703 and the information about the vehicle registered in steps S704 to 705. The discarding of the recognition information 500 may be, for example, deleting the recognition information 500 from the storage unit 410. In another embodiment, the discarding of the recognition information 500 may be performed by sending the recognition information 500 to the center device 2. It may not be transmitted. If the recognition information 500 is discarded in step S710, the flow returns to step S701.

  When the vehicle license plate symbol detected in the second lane 202 fails to be recognized by the operation flow of FIG. 7 described above, the adjacent first lane 201 in the predetermined period corresponding to the time when the vehicle is detected. Recognition information 500 obtained by imaging is obtained. Then, it is determined whether or not a vehicle whose vehicle type is determined to be large is detected in the recognition information 500 for the first lane 201 adjacent to the acquired predetermined period. When a vehicle whose type is determined to be large is detected, the failure to recognize the license plate symbol indicates that a large vehicle passing through the adjacent lane appears in the image of the second imaging device 212 in the second lane 202. There is a high possibility that this is due to Therefore, the recognition information 500 in which the image that failed to recognize the license plate symbol is registered is discarded. Therefore, the recognition information 500 resulting from a large vehicle passing through the adjacent lane appearing in the image of the second imaging device 212 in the second lane 202 is not collected by the center device 2. Therefore, it is possible to suppress erroneous detection of a vehicle traveling in the first lane 201 on the near side that is not the imaging target as a vehicle traveling on the second lane 202 that is the imaging target. Further, since the recognition information 500 including the image of the vehicle that failed to recognize the license plate symbol due to the erroneous detection is discarded, the recognition failure due to the erroneous detection fails to identify the license plate symbol. It will be excluded from the case that shows that. Accordingly, it is possible to prevent a decrease in the (successful recognition rate of license plate symbols).

  In the operation flow of FIG. 7 described above, in the process of step S704, the control unit 400 of the recognition device 3 functions as the specifying unit 411, for example. In the processing from step S706 to step S708, the control unit 400 of the recognition apparatus 3 functions as, for example, the determination unit 412. In the process of step S710, the control unit 400 of the recognition device 3 functions as, for example, the exclusion unit 413.

  Next, the second embodiment will be described. In the first embodiment described above, when a plurality of imaging devices 4 (for example, the first imaging device 211 and the second imaging device 212) are connected to the recognition device 3, for example, as in the installation position 1 of FIG. An example was given. In this case, the recognition device 3 executes the operation flow described in FIGS. 6 and 7 to detect the vehicle from a plurality of lanes (for example, the first lane 201 and the second lane 202), and the recognition information for each lane. 500 is generated and transmitted to the center apparatus 2. On the other hand, in the second embodiment, an example in which a set of the recognition device 3 and the imaging device 4 is installed for each lane to be imaged as in the installation position 2 of FIG.

  In 2nd Embodiment, the recognition apparatus 3 connected to the 1st imaging device 211 which images the 1st lane 201 performs the process similar to the operation | movement flow of FIG. The recognition information 500 that is the result of executing is transmitted to the center device 2.

  Further, the recognition device 3 connected to the second imaging device 212 that captures the second lane 202 may execute the same processing as the operation flow of FIG. 7 except for step S707. However, in step S707, the recognition device 3 connected to the second imaging device 212 notifies the recognition device 3 connected to the first imaging device 211 of the imaging time of the image in which the vehicle was detected in step S703. To do. Then, as a response to the notification of the imaging time, the recognition information 500 of the adjacent first lane 201 in the predetermined period specified based on the notified imaging time is recognized from the recognition device 3 connected to the first imaging device 211. Receive and get.

  When the recognition device 3 connected to the first imaging device 211 receives the notification of the imaging time from the recognition device 3 connected to the second imaging device 212 in step S707 of FIG. 7, the recognition information shown in FIG. The execution of the operation flow of 500 notification processing is started. The operation flow of the notification process of the recognition information 500 in FIG. 8 is such that, for example, the program 420 stored in the storage unit 410 of the recognition device 3 connected to the first imaging device 211 is connected to the first imaging device 211. This is implemented by reading and executing the control unit 400 of the recognition device 3. In step S801, the control unit 400 of the recognition device 3 connected to the first imaging device 211 images the first lane 201 with the first imaging device 211 during a predetermined period specified based on the notified time. The recognition information 500 generated in this way is extracted. Note that the predetermined period may be, for example, the same period as described in the description of step S707 in the first embodiment. In step S802, the control unit 400 of the recognition device 3 connected to the first imaging device 211 transmits the recognition information 500 extracted in step S801 to the recognition device 3 connected to the second imaging device 212.

  As described above, in the second embodiment, the recognition device 3 connected to the second imaging device 212 uses the recognition information 500 obtained by executing the vehicle recognition process on the first lane 201 as the first information. Obtained from the recognition device 3 connected to the imaging device 211. Therefore, as shown in the installation position 2 in FIG. 1, even when a set of the recognition device 3 and the imaging device 4 is installed for each lane to be imaged, a vehicle traveling in the first lane 201 on the near side that is not an imaging target is It is possible to suppress erroneous detection as a vehicle traveling in the second lane 202 to be imaged. In addition, it is possible to prevent a decrease in the (recognition success rate of the license plate symbol) due to the erroneous detection.

  Subsequently, a third embodiment will be described. In the third embodiment, it is determined whether or not the license plate recognition failure for the vehicle detected by the second imaging device 212 is caused by a large vehicle passing through the adjacent first lane 201. An example in which the center apparatus 2 executes will be described.

  FIG. 9 is a diagram illustrating a functional block configuration of the center apparatus 2 according to some embodiments including the third embodiment. The center device 2 includes a control unit 900 and a storage unit 910, for example. The control unit 900 includes functional units 901 such as a determination unit 911 and an exclusion unit 912, for example. The storage unit 910 of the center device 2 stores, for example, information such as a program 920 and recognition information 500 (for example, first recognition information 500 (1) and second recognition information 500 (2)). The control unit 900 of the center apparatus 2 functions as a function unit 901 such as a determination unit 911 and an exclusion unit 912 by reading and executing the program 920. Details of these functional units 901 and details of information stored in the storage unit 910 will be described later.

  In the third embodiment, each of the recognition device 3 connected to the first imaging device 211 and the recognition device 3 connected to the second imaging device 212 performs the same processing as the operation flow shown in FIG. As a result, the recognition information 500 is transmitted to the center device 2. When the center device 2 receives the recognition information 500 (hereinafter referred to as the first recognition information 500 (1)) transmitted from the recognition device 3 connected to the first imaging device 211, the center device 2 stores it in the storage unit 910. Store.

  When the center device 2 receives the recognition information 500 (hereinafter referred to as the second recognition information 500 (2)) transmitted from the recognition device 3 connected to the second imaging device 212, the center device 2 is shown in FIG. The execution of the operation flow of the exclusion process by the center device 2 is started. Note that the operation flow of FIG. 10 is implemented, for example, when the control unit 900 of the center apparatus 2 reads and executes the program 920 stored in the storage unit 910.

  In step S1001, the control unit 900 of the center apparatus 2 refers to, for example, the land transportation branch office code, classification number, usage code, serial designation number, etc. in the item 501 of the received second recognition information 500 (2) to recognize the license plate. It is determined whether or not successful. If the recognition has succeeded (YES in step S1001), the flow proceeds to step S1004. On the other hand, if the recognition has failed (NO in step S1001), the flow proceeds to step S1002. In step S1002, the control unit 900 of the center device 2 acquires the item 501: passage date and time of the second recognition information 500 (2) that has failed to be recognized in step S1001. And the 1st recognition information 500 (1) which has the item 501 within the predetermined period specified based on the passage date and time and the passage date and time is acquired from the memory | storage part 910. FIG. Note that the predetermined period may be, for example, the same period as described in the description of step S707 in the first embodiment. In step S1003, the control unit 900 of the center device 2 refers to the item 501 of the acquired first recognition information 500 (1): the value 502 of the vehicle type, and determines whether or not a large vehicle is included. If the vehicle type does not include a large vehicle (NO in step S1003), the flow proceeds to step S1004. In step S1004, the control unit 900 of the center apparatus 2 stores the received second recognition information 500 (2) in the storage unit 910, and the operation flow ends.

  On the other hand, if a vehicle with a large vehicle type is included in step S1003 (YES in step S1003), the flow proceeds to step S1005. In step S1005, the control unit 900 of the center device 2 excludes the received second recognition information 500 (2) from the case indicating that the identification of the license plate symbol has failed, and the operation flow ends. Note that the exclusion from the case indicating that the identification of the symbol in step S1005 has failed may be to delete the received data of the second recognition information 500 (2). In this case, since the received data of the second recognition information 500 (2) has been deleted from the storage unit 910, it is excluded from the example indicating that the identification of the license plate symbol has failed, and (the license plate symbol The recognition success rate is not used for calculation. Alternatively, the exclusion from the case indicating that the identification of the symbol in step S1005 has failed may be, for example, adding a flag indicating the exclusion to the received second recognition information 500 (2). In this case, for example, the second recognition information 500 (2) having an exclusion flag is configured not to be used in the calculation of the (recognition success rate of the license plate symbol), thereby failing to identify the license plate symbol. The second recognition information 500 is excluded from the case that indicates this. For example, by configuring as described above, it is possible to prevent a decrease in the (recognition success rate of license plate symbols) also in the third embodiment.

  In the operation flow of FIG. 10 described above, in the processing from step S1001 to step S1003, the control unit 900 of the center apparatus 2 functions as the determination unit 911, for example. In the process of step S1005, the control unit 900 of the center apparatus 2 functions as, for example, the excluding unit 912.

FIG. 11 is a diagram illustrating a hardware configuration of a computer 1100 for realizing the recognition apparatus 3 according to some embodiments. A hardware configuration for realizing the recognition device 3 in FIG. 11 includes, for example, a processor 1101, a memory 1102, a storage device 1103, a communication interface 1106, and an input / output device 1107. Note that the processor 1101, the memory 1102, the storage device 1103, the communication interface 1106, and the input / output device 1107 are connected to each other via a bus 1108, for example.
The processor 1101 provides a part or all of the functions of the functional units described above by executing a program describing the procedure of the operational flow described above using the memory 1102. The control unit 400 of the recognition device 3 is, for example, a processor 1101, and the storage unit 410 includes, for example, a memory 1102 and a storage device 1103. For example, the processor 1101 of the recognition device 3 functions as the specifying unit 411, the determination unit 412, and the exclusion unit 413 by reading and executing the program 420 stored in the storage device 1103. Also, for example, recognition information 500 is stored in the storage device 1103 of the recognition device 3.

The memory 1102 is a semiconductor memory, for example, and includes a RAM area and a ROM area. The storage device 1103 is, for example, a hard disk, a semiconductor memory such as a flash memory, or an external storage device.
The communication interface 1106 transmits and receives data via the network 1120 in accordance with instructions from the processor 1101. The input / output device 1107 is connected to a device such as an imaging device 1110, for example, and functions as an interface with the connected device. The above-described imaging device 4 is, for example, the imaging device 1110, and the imaging device 1110 includes, for example, a camera. The imaging device 1110 acquires, for example, an image obtained by imaging a lane to be imaged, and the acquired image is stored in the storage device 1103. In another example, an image acquired by the imaging device 1110 is transmitted to the center device 2 via the communication interface 1106. In another embodiment, the computer 1100 may acquire an image from the imaging device 1110 connected via the communication interface 1106.

The program according to the embodiment is provided to the recognition device 3 in the following form, for example.
(1) Installed in advance in the storage device 1103.
(2) Provided from a server 1130 such as a program server.

FIG. 12 is a diagram illustrating a hardware configuration of a computer 1200 for realizing the center device 2 according to some embodiments. The hardware configuration for realizing the center device 2 in FIG. 12 includes, for example, a processor 1201, a memory 1202, a storage device 1203, a reading device 1204, a communication interface 1206, and an input / output device 1207. Note that the processor 1201, the memory 1202, the storage device 1203, the reading device 1204, the communication interface 1206, and the input / output device 1207 are connected to each other via a bus 1208, for example.
The processor 1201 provides a part or all of the functions of the functional units described above by executing a program describing the procedure of the operational flow described above using the memory 1202. The control unit 900 of the center device 2 is, for example, a processor 1201, and the storage unit 910 includes, for example, a memory 1202, a storage device 1203, and a removable storage medium 1205. For example, the processor 1201 of the center device 2 functions as the determination unit 911 and the exclusion unit 912 by reading and executing the program 920 stored in the storage device 1203.

  The storage device 1203 of the center device 2 stores, for example, recognition information 500 (for example, first recognition information 500 (1) and second recognition information 500 (2)).

  The memory 1202 is a semiconductor memory, for example, and includes a RAM area and a ROM area. The storage device 1203 is, for example, a hard disk, a semiconductor memory such as a flash memory, or an external storage device.

  The reading device 1204 accesses the removable storage medium 1205 in accordance with instructions from the processor 1201. The removable storage medium 1205 is, for example, a semiconductor device (USB memory or the like), a medium (such as a magnetic disk) to which information is input / output by a magnetic action, or a medium (CD-ROM, etc.) to which information is input / output by an optical action. For example, a DVD). The communication interface 1206 transmits and receives data via the network 1220 in accordance with instructions from the processor 1201. The input / output device 1207 corresponds to, for example, an interface between a device that receives an instruction from a user such as a keyboard, a display device such as a display, and an audio device such as a speaker.

Each program according to the embodiment is provided to the center apparatus 2 in the following form, for example.
(1) Installed in advance in the storage device 1203.
(2) Provided by the removable storage medium 1205.
(3) Provided from a server 1230 such as a program server.

  In the above, several embodiments have been described. However, the embodiments are not limited to the above-described embodiments, and should be understood as including various modifications and alternatives of the above-described embodiments. For example, it will be understood that various embodiments can be embodied by modifying the components without departing from the spirit and scope thereof. It will be understood that various embodiments can be made by appropriately combining a plurality of components disclosed in the above-described embodiments. Alternatively, various embodiments may be implemented by deleting or replacing some components from all the components shown in the embodiments, or adding some components to the components shown in the embodiments. Those skilled in the art will understand that

DESCRIPTION OF SYMBOLS 1 System 2 Center apparatus 3 Recognition apparatus 4 Imaging device 5 Network 400 Control part 410 Memory | storage part 411 Specification part 412 Determination part 413 Exclusion part 900 Control part 910 Storage part 911 Determination part 912 Exclusion part 1100 Computer 1101 Processor 1102 Memory 1103 Storage apparatus 1106 Communication interface 1107 Input / output device 1108 Bus 1110 Imaging device 1120 Network 1130 Server 1200 Computer 1201 Processor 1202 Memory 1203 Storage device 1204 Reading device 1205 Removable storage medium 1206 Communication interface 1207 Input / output device 1208 Bus 1220 Network

Claims (6)

The second lane installed at a position farther away from the first lane and the second lane adjacent to the first lane than the distance to the first lane that is the imaging target of the first imaging device. A specifying unit for specifying a symbol written on a license plate of the vehicle from an image of the vehicle imaged by the second imaging device to be imaged;
When the identification unit fails to identify the symbol, a predetermined condition is satisfied from the image captured by the first imaging device during a predetermined period specified based on the time when the image of the vehicle is captured. A determination unit for determining whether or not a vehicle is detected;
If the determination unit determines that the vehicle has been detected, an exclusion unit that excludes the vehicle image captured by the second imaging device from an example indicating that the identification of the license plate symbol has failed;
A recognition device comprising:   The recognition apparatus according to claim 1, wherein the predetermined condition is that a license plate of a vehicle detected from an image captured by the first imaging device is determined as a large number mark.   2. The predetermined condition is that the vehicle width, height, area, or depth detected from an image captured by the first imaging device is determined to be equal to or greater than a predetermined value. Recognition device. The second lane installed at a position farther away from the first lane and the second lane adjacent to the first lane than the distance to the first lane that is the imaging target of the first imaging device. From the image of the vehicle imaged by the second imaging device to be imaged, the symbol described on the license plate of the vehicle is identified,
When the identification of the symbol fails, a vehicle satisfying a predetermined condition is detected from the image captured by the first imaging device during a predetermined period specified based on the time when the image of the vehicle is captured. Whether or not
When it is determined that the vehicle has been detected, the vehicle image captured by the second imaging device is excluded from an example indicating that the identification of the license plate symbol has failed,
A program that causes a computer to execute processing. The second lane installed at a position farther away from the first lane and the second lane adjacent to the first lane than the distance to the first lane that is the imaging target of the first imaging device. Identifying a symbol described on a license plate of the vehicle from an image of the vehicle imaged by the second imaging device to be imaged;
If the identifying step fails to identify the symbol, a predetermined condition is determined from the image captured by the first imaging device during a predetermined period specified based on the time when the image of the vehicle is captured. Determining whether a satisfying vehicle has been detected; and
If the step of determining determines that the vehicle has been detected, the step of excluding the image of the vehicle imaged by the second imaging device from an example indicating that identification of a license plate symbol has failed;
A method performed by a computer, including: Of the images captured by the first imaging device that captures the first lane, the first image in which the vehicle is detected, the imaging time of the first image, and the first image. A first recognition device that registers the vehicle type determined from the vehicle in the first recognition information, and transmits the first recognition information to the center device;
A second imaging device that targets the second lane and is located at a position farther away from the first lane than to the second lane adjacent to the first lane. Among the images captured in step 2, the second image in which the vehicle is detected, the image capture time of the second image, and the symbols described on the license plate executed for the vehicle in the second image A second recognition device that registers information indicating success or failure of the recognition in the second recognition information, and transmits the second recognition information to the center device;
When receiving second recognition information including information indicating that the recognition of the symbol described in the license plate has failed from the second recognition device, the identification is performed based on the imaging time included in the second recognition information. It is determined whether or not the first recognition information including the imaging time of the first image having the imaging time of the predetermined period is registered with the first recognition information in which the vehicle type determined to be large is registered. Judgment part,
When there is the first recognition information in which the vehicle type determined to be large is registered, the second recognition information including information indicating that the recognition of the symbol described in the received license plate has failed is performed. Exclusion part to exclude from cases showing failure to identify the symbol,
Including the center device,
Including system.