JP2006202000A - Method for monitoring vehicle, apparatus for recognizing particular character pattern, and vehicle monitoring system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To control the amount of transfer data according to the result of recognition of a particular character pattern, relating to a method for monitoring a vehicle through the recognition of the vehicle's particular character pattern, an apparatus for recognizing the particular character pattern, and a vehicle monitoring system. <P>SOLUTION: The vehicle monitoring system includes recognition devices 7 for recognizing the particular character pattern of a vehicle by processing image data created by an imaging part, and a center 1 connected to the plurality of recognition devices via a line 6. The recognition device 7 implements a process for recognizing the vehicle's particular character pattern on the basis of the image data; if no character pattern can be recognized, the recognition device cuts a vehicle region out of the image data and compresses the region at a low compression ratio; if only a portion of the character pattern can be recognized, the recognition device cuts out the region of the particular character pattern and compresses the region at a low compression ratio; if the entire character pattern can be recognized, the recognition device cuts out the particular character pattern, compresses the pattern at a high compression ratio, and either records or transmits the pattern. The center 1 has a management table 4 for receiving and storing compressed data from the recognition device via the line 6, and a display part 3 for displaying the data. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a vehicle monitoring method, a specific character pattern recognition device, and a vehicle monitoring system that recognize and monitor a specific character pattern of a vehicle.

  Specific character patterns are assigned to various vehicles so as not to overlap each other. Therefore, by automatically reading the specific character pattern and recognizing the vehicle specific character pattern, it is possible to automate the toll road toll booth, determine whether the parking vehicle in the parking lot belongs to the contractor or theft Various systems have already been proposed, which can be applied to detection of vehicles and escape vehicles. In this case, an image signal obtained by capturing a specific character pattern is processed. Therefore, it is desired to efficiently process the image signal.

  FIG. 7 is an explanatory diagram of the main part of the conventional example. The vehicle detection unit 70 detects that the passing vehicle has reached the shooting position, and inputs the detection signal to the recognition processing unit 72. The recognition processing unit 72 instructs the shooting unit 71 to capture a passing vehicle. The photographing unit 71 photographs the vehicle according to the photographing instruction, and transfers the image signal to the recognition processing unit 72 and the image transmission unit 74. The recognition processing unit 72 extracts a specific character pattern area in the photographed vehicle image, and recognizes a vehicle specific character pattern using characters and numbers in the specific character pattern area. For the recognition processing of the vehicle specific character pattern, various means such as extraction of one character, pattern matching for the extracted character, and collation by feature extraction are known. In this conventional example, the vehicle is photographed by the photographing unit 71 only when the vehicle passes the photographing position.

  The collation unit 73 registers a specific vehicle specific character pattern in advance, transfers the vehicle specific character pattern recognized by the recognition processing unit 72 to the collation unit 73, and collates with the registered vehicle specific character pattern. The collation result is transferred to the recognition processing unit 72. When receiving the collation match signal from the collation unit 73, the recognition processing unit 72 transfers the collated vehicle specific character pattern to the image transmission unit 74. The image transmission unit 74 transfers the vehicle specific character pattern and the captured image signal to the image recording / display unit 75. The image recording / display unit 75 adds and records date and time information and shooting point information, and if necessary, searches and reads based on information such as the vehicle specific character pattern, date and time, shooting point, and displays it on the monitor. Thus, visual confirmation is possible (for example, see Patent Document 1).

  The vehicle specific character pattern on the specific character pattern is recognized from the captured image signal, the captured image signal is compressed, and the recognized vehicle specific character pattern is added and recorded. There is also known a means for reading out the recorded content based on a vehicle specific character pattern, decompressing the read compressed image signal, and displaying it on a monitor (for example, see Patent Document 2).

There is also known remote monitoring means for selecting one monitoring range from among a plurality of monitoring ranges and switching the image signal compression rate and transmission time in accordance with the monitoring range. For example, in the case of a selected monitoring range in which the monitored object moves at high speed, the compression rate is increased and high-speed transmission is performed so that the monitor screen changes following the monitored object moving at high speed. In the case of a selected monitoring range where the monitored object moves at a low speed, the compression rate is lowered and the transmission is performed at a low speed, and the resolution of the monitor screen is increased (for example, see Patent Document 3).
JP-A-9-282590 JP-A-9-261620 JP-A-9-37232

  As described above, it is possible to capture a character pattern that identifies a vehicle, extract the specific character pattern by image processing, perform character recognition processing, and perform vehicle monitoring based on the recognized character pattern. However, the shooting conditions differ greatly depending on the angle of attachment of the specific character pattern, the state of dirt, and the weather. Therefore, in an unfavorable condition, it is difficult to recognize a character pattern by image processing. When such automatic recognition processing of the specific character pattern of the vehicle is difficult, the captured image signal is input and displayed on the monitor, and the region of the specific character pattern is visually identified based on the display screen. A means for recognizing the pattern can be employed.

  However, when the monitor installation location and, for example, the image transmission unit 74 shown in FIG. 7 are installed away from each other, the image data is transmitted via the line, but the image data is compressed. However, since the amount of data for one screen is large, a high-speed transmission path is required. Therefore, it has been difficult to remotely monitor the vehicle by connecting between the center and the imaging processing units arranged in a distributed manner via a low-rate network.

  An object of the present invention is to reduce transmission data by controlling a compression range and a compression rate according to a recognition result of a specific character pattern of a vehicle.

  The vehicle monitoring method of the present invention is a method for extracting a specific character pattern area of a vehicle based on captured image data, performing character pattern recognition processing, and performing vehicle monitoring based on the recognized character pattern, When the result of the character pattern recognition process indicates that the character pattern cannot be recognized, a vehicle area including almost the entire vehicle is cut out from the captured image data, compressed at a low compression rate, and recorded or transmitted. Further, when it is shown that only a part of the vehicle specific character pattern can be recognized, the specific character pattern region is cut out from the captured image data, compressed at a low compression rate, and recorded or transmitted. Further, when the fact that all the vehicle specific character patterns have been recognized, the specific character pattern area is cut out from the captured image data, compressed at a high compression rate, and recorded or transmitted. Therefore, when all the vehicle specific character patterns are recognized, the specific character pattern area having a relatively small area is transmitted as image data compressed at a high compression rate, so that it can be easily transmitted even on a low rate line. .

  The specific character pattern recognition apparatus of the present invention also includes a frame memory for temporarily storing captured image data by the image capturing unit, and an image process for extracting a vehicle area including a vehicle from the captured image data and extracting the specific character pattern area from the vehicle area. A recognition processing unit including a function and a character recognition processing function for recognizing a character pattern in a specific character pattern area, and when the result of the character pattern recognition processing by the recognition processing unit indicates that the character pattern cannot be recognized, When the compression range is the vehicle area, the compression ratio is the low compression ratio, and only a part of the character pattern is recognized, the compression range is the specific character pattern area, the compression ratio is the low compression ratio, and the entire character pattern is A compression range / compression rate determination unit that determines that the compression range is a specific character pattern area and the compression rate is a high compression rate when indicating that it has been recognized, and this compression In accordance with the determination result of the surrounding / compression rate determination unit, an area in the compression range is extracted from the frame memory, and compression processing is performed in accordance with the compression rate, and image data compressed by the image compression unit is transmitted. And a transmission / reception unit.

  A data storage unit that stores the image data compressed by the image compression unit together with the recognition result of the vehicle specific character pattern can be provided.

  In addition, the vehicle specific character pattern instructed to be searched is compared with the vehicle specific character pattern of the recognition result stored in the data storage unit, and the data stored in the data storage unit is read out and transmitted when the collation matches. A verification processing unit can be provided.

  The vehicle monitoring system according to the present invention includes a recognition device 7 that recognizes a vehicle-specific character pattern by processing image data captured by a photographing unit, and a center 1 that is connected to a plurality of recognition devices via a line 6. The recognition device 7 performs recognition processing of a vehicle specific character pattern based on photographed image data. When this character pattern cannot be recognized, a vehicle region is cut out from the photographed image data and compressed at a low compression rate. When only a part of the character pattern is recognized, the specific character pattern area is cut out and compressed at a low compression rate. When the entire character pattern is recognized, the specific character pattern area is cut out and compressed at a high compression rate. The center 1 has a management table 4 for receiving and storing compressed data from the recognition apparatus via the line 6 and a display unit 3 for displaying the compressed data. It is.

  The area and compression ratio of the captured image data are selected according to the recognition result in the automatic recognition of the vehicle specific character pattern. Even if the vehicle specific character pattern is not recognized, the vehicle area is compressed instead of the full screen. Since processing is performed, the amount of data can be reduced. In addition, since the vehicle portion can be displayed with high image quality when displayed on the display unit 3 of the center 1, the possibility of recognizing the vehicle specific character pattern can be increased by visual observation. When the vehicle specific character pattern is recognized, the specific character pattern area with a smaller area is compressed, so that the amount of data can be further reduced and can be easily transmitted via a low-rate line. Is possible. In addition, since the recognition rate of the vehicle specific character pattern is relatively high, even when continuously transmitting compressed data, most of the data is highly compressed data in the specific character pattern area, and the recognized vehicle specific character pattern data etc. Even if the transmission line is included, since the amount of transmission data is small, it can be easily transmitted even on a low-speed line.

  According to the vehicle monitoring method of the present invention, when the result of the recognition process of the vehicle specific character pattern indicates that the character pattern cannot be recognized, the vehicle region including almost the entire vehicle is cut out from the captured image data and compressed at a low compression rate. Record or transmit. Further, when it is shown that only a part of the vehicle specific character pattern can be recognized, the specific character pattern region is cut out from the captured image data, compressed at a low compression rate, and recorded or transmitted. Further, when the fact that all the vehicle specific character patterns have been recognized, the specific character pattern area is cut out from the captured image data, compressed at a high compression rate, and recorded or transmitted.

  A specific character pattern recognition device according to the present invention includes a frame memory that temporarily stores image data captured by an imaging unit, and an image processing function that extracts a vehicle area including a vehicle from the captured image data and extracts a specific character pattern area from the vehicle area. And a recognition processing unit including a character recognition processing function for recognizing a character pattern in a specific character pattern area, and the result of the character pattern recognition processing by the recognition processing unit indicates that the character pattern cannot be recognized. When the range is a vehicle area, the compression ratio is a low compression ratio, and when only a part of a character pattern is recognized, the compression range is a specific character pattern area, the compression ratio is a low compression ratio, or A compression range / pressure that indicates that the compression range is the specific character pattern area and the compression ratio is a high compression ratio when indicating that the entire character pattern has been recognized. A rate determination unit, an image compression unit that extracts a compression range area from the frame memory in accordance with a determination result of the compression range / compression rate determination unit, and performs compression processing according to the compression rate, and the image compression unit A transmission / reception unit for transmitting compressed image data.

  The vehicle monitoring system of the present invention includes a recognition device 7 that recognizes a vehicle-specific character pattern by processing captured image data by a photographing unit, and a center 1 that is connected to a plurality of recognition devices via a line 6. 7 performs recognition processing of the vehicle specific character pattern based on the photographed image data. When this character pattern cannot be recognized, the vehicle region is cut out from the photographed image data and compressed at a low compression rate, and only a part of the character pattern is recognized. When it is recognized, the specific character pattern area is cut out and compressed at a low compression rate. When the entire character pattern is recognized, the specific character pattern area is cut out and compressed at a high compression rate, and recorded or transmitted. The center 1 includes a management table 4 that receives and stores compressed data from the recognition apparatus via the line 6 and a display unit 3 that displays the management table 4.

  FIG. 1 is a schematic explanatory diagram of Embodiment 1 of the present invention, where 1 is a center, 2 is a control processing unit, 3 is a display unit, 4 is a data management table, 5 is a switching network (network), 6 is a telephone line, etc. , 7 is a recognition device, 8 is a camera, and 9 is a vehicle. A case where the vehicle monitoring system is configured by connecting the center 1 and the recognition device 7 via a line 6 accommodated in the exchange network 5 is shown. In addition, the recognition device 7 can be distributed on an expressway, a general road, a parking lot, or the like and connected to the center 10 via a line to constitute a vehicle monitoring system.

  The camera 8 is disposed at a position for photographing a vehicle traveling on a highway or a general road, or a position for photographing a vehicle entering or exiting a parking lot or a parked vehicle. The recognition device 7 has an image processing function, a character pattern recognition processing function, a data transmission / reception function, and the like. The image data from the camera 8 is processed by the image processing function to extract a specific character pattern area of the vehicle 9. The vehicle specific character pattern is recognized by the character recognition processing function, and the region including the entire vehicle 9 or the specific character pattern region is cut out from the captured image data corresponding to the recognition processing result in this case, and the compression rate is selected. The data is sent to the center 1 via the line 6 by the transmission / reception function.

  The transmission data in this case includes specific character pattern information and compressed image data, as shown in the figure. For example, as specific character pattern information, header information, point information, time information, vehicle-specific character patterns of recognition results, for example, “12345” is included, and the compressed image data includes a compression range of a vehicle area or a specific character pattern area, an image quality, a compressed data length, and compressed data.

  The control processing unit 2 of the center 1 includes a data management table 4 which stores transmission data transmitted from the recognition device 7 via the line 6 and is partially omitted. Vehicle-specific character patterns, image data within the compression range, image quality, and time.

  For example, “12345” in the column of the specific character pattern of item number 1 indicates a case where all the vehicle specific character patterns can be recognized, and the image data indicates a plate portion and the image quality is low. That is, the compression range is a specific character pattern area, and the compression rate is increased so that the reproduction image quality may be low. In addition, “?” In the specific character pattern column of item number 3 indicates a case where all the vehicle specific character patterns are not recognized. In this case, the image data is the entire vehicle and the image quality is high. That is, the compression range is set to the vehicle area, and the compression rate is lowered so that the reproduction image quality is improved. “12? 45” in the column of the specific character pattern of item number 4 indicates that the character “?” Of the vehicle specific character pattern is not recognized, and that the image data has a specific character pattern portion and the image quality is high. That is, the compression range is set to a specific character pattern region, and the compression rate is lowered so that the reproduction image quality is improved.

  For example, when the image data of item No. 3 in the data management table 4 is read and displayed on the display unit 3 under the control of the control processing unit 2, the compressed data of the low compression ratio vehicle area is restored and displayed. The vehicle can be displayed with image quality. Therefore, the specific character pattern area can be confirmed visually on the display screen, and the characters in the specific character pattern area can be read. In that case, the column of the specific character pattern can be corrected to the vehicle specific character pattern that can be visually confirmed. Compared with the case where the entire screen is compressed, only the vehicle area is compressed, so that the amount of data can be reduced.

  When the image data of item number 4 is read out and displayed on the display unit 3, the specific character pattern portion can be displayed with high image quality. Therefore, it becomes easy to visually confirm the character “?” That could not be automatically recognized. Also in this case, the specific character pattern column of the data management table 4 can be corrected according to the confirmed contents. In this case, since only the specific character pattern area is compressed and transmitted, data can be reduced as compared with the case where the entire screen or the vehicle area is compressed.

  If the vehicle specific character pattern can be automatically recognized as in item Nos. 1, 2, and 6, the specific character pattern area is compressed so that the image quality may be low, so the data amount can be further reduced. Can do. In this case, when a specific character pattern that has already been automatically recognized is displayed on the display unit 3, reconfirmation is easy even if the image quality is low.

  FIG. 2 is an explanatory diagram of a system according to the first embodiment of the present invention. 10 is a center, 11 is a transmission / reception unit, 12 is a control processing unit, 13 is a display unit, 14 is a management table, 20 is a specific character pattern recognition device, 21 Is a recognition processing unit, 22 is a frame memory, 23 is a compression range / compression ratio determination unit, 24 is an image compression unit, 25 is a transmission / reception unit, 30 is a photographing unit, and 40 is a line such as a telephone line. The center 10 corresponds to the center 1 in FIG. 1, the specific character pattern recognition device 20 corresponds to the recognition device 7 in FIG. 1, and the photographing unit 30 corresponds to the camera 8 in FIG. The line 40 corresponds to the line 6 in FIG. The control processing unit 12 corresponds to the control processing unit 2 in FIG. 1, the display unit 13 corresponds to the display unit 3 in FIG. 1, and the management table 14 is the data management table in FIG. Corresponding to 4. The network 5 in FIG. 1 is not shown.

  The photographing unit 30 can continuously set the passing position or the parking position of the vehicle as a photographing range in a predetermined frame cycle as in the television camera. In contrast to this, it is of course possible to take a picture only when the vehicle is within the shooting range. The image signal from the photographing unit 30 is digitized and stored in the frame memory 22 of the recognition processing unit 21. As described above, when continuously capturing images, the recognition processing unit 21 can distinguish between a background portion such as a road and a moving vehicle portion by a difference between frames at a predetermined time interval. An area including the vehicle can be cut out. And the specific character pattern area in the vehicle area is cut out, and the character recognition process in the specific character pattern area is performed.

  The compression range / compression rate determination unit 23 controls the compression range and the compression rate according to the recognition processing result of the vehicle specific character pattern in the recognition processing unit 21. As described above, when the vehicle specific character pattern cannot be recognized, the compression range is set as the vehicle region, and the compression rate is lowered. When all the vehicle specific character patterns are recognized, the compression range is set as the specific character pattern region, and the compression ratio is increased. If a part of the vehicle specific character pattern can be recognized and the rest cannot be recognized, the compression range is set as the specific character pattern region, and the compression rate is lowered.

  The image compression unit 24 performs image data area extraction and compression processing in the frame memory 22 according to the determination signal from the compression range / compression rate determination unit 23, and converts the compressed image data into the compression range / compression rate determination unit 23. Or directly to the transmitting / receiving unit 25 and transmitted to the center 10 via the line 40.

  In the center 10, the compressed image data received by the transmission / reception unit 11 is temporarily stored in the management table 14 under the control of the control processing unit 12, and a vehicle whose vehicle specific character pattern cannot be automatically recognized is automatically detected. Alternatively, it is read from the management table 14 according to an operator's instruction, decompressed in accordance with the compression rate, and displayed on the display unit 13. In this case, since the vehicle area can be displayed with high image quality, there is a high possibility that the vehicle specific character pattern can be visually recognized on the display screen.

  FIG. 3 is a flowchart of the first embodiment of the present invention. The captured image signal from the imaging unit 30 (see FIG. 2) is digitized and written to the frame memory 22 (A1), and the background image data, current captured image data, Difference data (vehicle contour) is extracted (A2). As described above, the difference data indicates a difference between a background screen that is the same screen even when time passes, such as a road, and a screen that is different from a background screen such as a vehicle. Accordingly, the difference data includes the moved vehicle. From the difference data, a region including the vehicle is cut out (A3).

  Then, it is determined whether or not the vehicle has been recognized (A4). That is, the vehicle has various shapes, but has a size and shape different from those of pedestrians and the like, so that the vehicle can be distinguished from other moving objects. If the vehicle cannot be recognized from the difference data, the process proceeds to step (A1). If the vehicle can be recognized, the process proceeds to step (A5).

  In step (A5), a specific character pattern area is extracted. For example, a histogram for the horizontal direction and the vertical direction is created, and a rectangular specific character pattern area is extracted based on the histogram. can do. Alternatively, if the background color of the specific character pattern is, for example, white, an area that is substantially rectangular and has a white background color other than characters can be extracted as the area of the specific character pattern.

  Then, it is determined whether or not the specific character pattern has been recognized (A6). If it can be determined, the character is cut out and the specific character pattern is digitized (A7). That is, character recognition is performed on the extracted character region by pattern matching, feature point extraction, or the like. Next, it is determined whether or not all the specific character patterns have been digitized, that is, whether or not all the vehicle specific character patterns have been recognized (A8). The area is compressed with a high compression ratio, that is, with a high compression ratio (A9).

  If all the specific character patterns cannot be digitized, that is, if only a part of the vehicle specific character pattern can be recognized, the recognized specific character pattern is digitized (A12), and the extracted specific character pattern area Is compressed at a low compression rate (A13). If the specific character pattern cannot be recognized, or if the vehicle specific character pattern cannot be recognized at all, the area including the cut-out vehicle is compressed with an emphasis on image quality, that is, with a low compression rate (A11). Then, transmission data as shown in FIG. 1 is created, and the compressed image data is transmitted from the specific character pattern recognition device 20 to the center 10 (A10).

  4A and 4B are explanatory diagrams of the compression range. FIG. 4A shows a photographed image by the photographing unit 30 (see FIG. 2), for example, a screen based on image data stored in the frame memory 22. When the specific character pattern cannot be extracted or when the vehicle specific character pattern cannot be automatically recognized, the vehicle area is cut out as shown in FIG. And compress at a low compression rate.

  Accordingly, when the compressed image data in this case is received at the center 10, restored and displayed on the display unit 13, a vehicle image with good image quality is displayed, and the specific character pattern region is confirmed visually. The possibility of recognizing the vehicle specific character pattern can be increased.

  When only a part of the vehicle specific character pattern can be automatically recognized, the specific character pattern area is compressed at a low compression rate. Accordingly, when image data obtained by compressing the specific character pattern area shown in (c) or (d) at a low compression rate is received by the center 10 and restored and displayed on the display unit 13, the specific character pattern area with good image quality is displayed. Since it becomes a display screen, the possibility of recognition of the vehicle specific character pattern by visual observation can be made high.

  FIG. 5 is a system explanatory diagram of Embodiment 2 of the present invention. The same reference numerals as those in FIG. The data storage unit 26 has a configuration for storing image data such as a semiconductor integrated circuit memory or a hard disk device. In the image compression unit 24, compressed data and recognition processing based on the determination result of the compression range / compression rate determination unit 23 The recognition result of the vehicle specific character pattern in the unit 21 is accumulated. For example, the compressed data is stored in the data storage unit 26 in the format of the transmission data shown in FIG.

  Then, the data storage unit 26 is searched from the center 10 in accordance with time specification, vehicle specific character pattern specification, etc., and the compressed data that has been searched and extracted is read out, for example, in the transmission data format shown in FIG. Send it out. Therefore, in the second embodiment, the vehicle specific character pattern and the compressed image data that are automatically recognized can be collected from the specific character pattern recognition device 20 only when necessary at the center 10.

  FIG. 6 is a system explanatory diagram of Embodiment 3 of the present invention. The same reference numerals as those in FIG. 2 or 5 indicate the same parts, and 27 is a collation processing unit. When a specific vehicle specific character pattern is specified from the center 10, the verification processing unit 27 searches for the vehicle specific character pattern stored in the data storage unit 26, and compares the specified vehicle specific character pattern with the recognized vehicle specific character pattern. In the case of collation matching, the compressed image data is read from the data storage unit 26 in accordance with, for example, the format of the transmission data shown in FIG. 1 and transmitted from the transmission / reception unit 25 to the center 10 via the line 6.

  The present invention is not limited to the above-described embodiments, and various additions and modifications can be made. For example, the camera 8 or the photographing unit 30 is a digital camera and photographs at a constant time interval. It is also possible. It is also possible to facilitate the processing when the difference data is obtained with the frame memory 22 having a two-plane configuration. It is also possible to control the compression rate so that it can be switched in multiple stages corresponding to the number of recognized vehicle specific character patterns.

It is outline | summary explanatory drawing of Example 1 of this invention. It is system explanatory drawing of Example 1 of this invention. It is a flowchart of Example 1 of this invention. It is explanatory drawing of a compression range. It is system explanatory drawing of Example 2 of this invention. It is system explanatory drawing of Example 3 of this invention. It is principal part explanatory drawing of a prior art example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Center 2 Control processing part 3 Display part 4 Data management table 5 Switching network 6 Line 7 Specific character pattern recognition apparatus 8 Camera 9 Vehicle

Claims (5)

In a method of extracting a specific character pattern area of a vehicle based on captured image data and performing a character pattern recognition process, and monitoring a vehicle based on the recognized character pattern,
When the result of the character pattern recognition process indicates that the character pattern of the vehicle cannot be recognized, a vehicle region including almost the entire vehicle is cut out from the captured image data, compressed at a low compression rate, and recorded or transmitted. When indicating that only a part of the specific character pattern of the vehicle can be recognized, the specific character pattern region is cut out from the captured image data, compressed at a low compression rate, recorded or transmitted, and the specific character of the vehicle A vehicle monitoring method comprising a step of cutting out a specific character pattern area from the photographed image data, compressing it with a high compression ratio, and recording or transmitting it when it indicates that all of the patterns have been recognized. A frame memory for temporarily storing image data captured by the imaging unit, an image processing function for extracting a vehicle area including a vehicle from the imaged image data, and extracting a specific character pattern area from the vehicle area, and characters in the specific character pattern area A recognition processing unit including a character recognition processing function for recognizing a pattern;
When the result of the character pattern recognition processing by the recognition processing unit indicates that the character pattern cannot be recognized, the compression range is the vehicle area, the compression rate is the low compression rate, and only a part of the character pattern is recognized. The compression range is set to a specific character pattern area, the compression rate is set to a low compression rate, and the compression range is determined to be a specific character pattern region and the compression rate is set to a high compression rate when the entire character pattern is recognized. A range / compression rate determination unit;
In accordance with the determination result of the compression range / compression rate determination unit, an image compression unit that extracts a compression range region from the frame memory and performs compression processing according to the compression rate;
A specific character pattern recognition apparatus comprising: a transmission / reception unit that transmits image data compressed by the image compression unit.   The specific character pattern recognition apparatus according to claim 2, further comprising a data storage unit that stores the image data compressed by the image compression unit together with a recognition result of the vehicle specific character pattern.   The vehicle specific character pattern instructed to be searched is compared with the vehicle specific character pattern of the recognition result stored in the data storage unit, and the data stored in the data storage unit is read out and sent out when the collation matches. 4. The specific character pattern recognition apparatus according to claim 3, further comprising a collation processing unit. In a vehicle monitoring system including a specific character pattern recognition device that recognizes a vehicle specific character pattern by processing captured image data by a photographing unit, and a center connected to a plurality of specific character pattern recognition devices via a line,
The specific character pattern recognition device performs a character pattern recognition process based on the captured image data. When the specific character pattern of the vehicle cannot be recognized, a vehicle region is extracted from the captured image data and compressed at a low compression rate. When only a part of the character pattern is recognized, the specific character pattern area is cut out and compressed with a low compression rate.When the whole character pattern is recognized, the specific character pattern area is cut out with a high compression rate. The center has a configuration for compressing and recording or sending, and the center has a management table for receiving and storing compressed data from the specific character pattern recognition device via a line, and a display unit for displaying the management table. Vehicle monitoring system.

Images