JP3275475B2 - Character string recognition device with known character sequence - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for recognizing a character string having a known character arrangement in an input image.

[0002]

2. Description of the Related Art Conventional character string recognition processing in which a character arrangement is known is performed by labeling processing in which image data around a character string is binarized and connected characters are extracted on a binarized image, or in vertical and horizontal directions. The character extraction position is determined by a method using the projection length data, and character recognition processing is performed from the result obtained.

[0003] In such a method, however, correct results can be obtained when the input image is good and stable.
It has been difficult to adequately cope with cases where the separation between characters is not clear due to dirt, shading, and binarization errors.

[0004]

DISCLOSURE OF THE INVENTION The present invention can be applied to a target character string whose character sequence is known, even if the input conditions fluctuate outdoors, changes around the character string, dirt, blur, or binarization error occur. An object of the present invention is to provide a character recognition device capable of extracting characters.

A first character recognition device according to the present invention comprises: input means for inputting an image including a target character string whose character arrangement is known; binarizing means for binarizing the input image;
The image data binarized by the binarizing unit is scanned with a plurality of types of extraction windows of different sizes each having an arrangement corresponding to a known character arrangement, and an extraction window is extracted for each of the plurality of types of extraction windows. A feature value calculating means for calculating a feature value for each position, a cutout position determining means for determining a character cutout position based on the feature value obtained by the feature value calculating means, and a position of the position determined by the cutout position determining means. Character recognition means for performing character recognition processing on image data cut out according to the extraction window based on a reference pattern stored in advance, and output means for outputting a result obtained by the character recognition means. is there.

[0006]

According to the first aspect of the present invention, instead of extracting a cutout position for each character as a cutout position determining means, a known character arrangement and size are used to extract a character arrangement corresponding to the character arrangement. By scanning the window, the character cutout position in the character string is determined. Therefore, environmental changes outdoors, changes around the target character string, dirt,
Characters can be correctly cut out even when the image is blurred, binarization error, or separation between characters is not clear, and correct character recognition processing can be performed.

[0008]

[0009]

[0010]

Further, when the CPU scans the extraction window, the CPU scans a plurality of extraction windows having different sizes corresponding to a plurality of character arrays assumed in advance, and obtains a character array corresponding to the character array from the obtained feature amounts. Determine the cutting position. Therefore, even in a target character string in which a plurality of character arrays having different sizes such as license plates exist, the corresponding character array can be correctly determined and the correct character recognition process can be performed.

[0012]

[0013]

[0014]

[0015]

According to a second character recognition device of the present invention, there is provided an input means for inputting an image including a target character string whose character arrangement is known, a binarizing means for binarizing the input image,
For image data binarized by the binarizing means, a plurality of types of extraction windows of different sizes having arrangements corresponding to known character arrangements are switched according to the position on the image, and scanning is performed. A feature value calculating means for calculating a feature value for each window position, a cutout position determining means for determining a character cutout position based on the feature value obtained by the feature value calculating means, and a position determined by the cutout position determining means Character recognition means for performing character recognition processing based on a reference pattern stored in advance for image data cut out according to the extraction window of (1), and output means for outputting a result obtained by the character recognition means. Things.

According to the second aspect of the present invention, instead of extracting the cutout position for each character as the cutout position determining means, the arrangement and size of a known character array are used to extract the cutout position corresponding to the character array. By scanning the window, the character cutout position in the character string is determined. Therefore, environmental changes outdoors, changes around the target character string, dirt,
Characters can be correctly cut out even when the image is blurred, binarization error, or separation between characters is not clear, and correct character recognition processing can be performed. In addition, a plurality of extraction windows having different sizes are prepared in advance, and the CP
When U scans the extraction window, the extraction window to be used is switched according to the scanning position on the image.
If the size of the characters in the image changes depending on the timing at which the input image is taken because the target character string is moving in a known fixed direction with respect to the image input means, for example, when the approaching vehicle Correct character recognition processing can also be performed on things such as license plates.

[0018]

[0019]

[0020]

[0021]

[0022]

[0023]

[0024]

[0025]

[0026]

FIG. 1 is an arrangement diagram of an embodiment in which a character recognition device is applied to read a license plate of a vehicle running on a road.

The pillars 3 are erected on both sides of the road 4, and the horizontal support 2 is passed over these columns, and the television (video) camera 1 is installed on the horizontal support 2. The camera 1 captures an image of a license plate of a vehicle 5 traveling in a direction toward the camera 1 as indicated by an arrow in a predetermined range on a road 4, and outputs a video signal representing the captured image. The video signal output from the camera 1 is transmitted to the I / O control device 11
To a character recognition device described later. The column 3 may be erected on either side of the road 4 and the camera 1 may be installed above the column. Further, the camera 1 can be attached to an overpass or other building to take an image of a license plate of a vehicle traveling on the road 4.

FIG. 2 is a view of the arrangement shown in FIG.

As shown in FIG. 2, since the visual angle θ of the camera 1 installed on the horizontal ridge 2 passed on the road 4 is set to be relatively narrow, the distance from the camera 1 to the recognition target in one image is set. There is almost no difference in size depending on the location depending on the distance, and the image of the license plate 6 is not distorted.

This character recognition device can be widely applied to other than a vehicle license plate as long as the arrangement of the target character strings is known. For example, reading the IC on the printed circuit board, the model number of the product flowing on the line, the number of the railcar, the characters and symbols indicating the delivery destination displayed on the delivery goods in the distribution industry, and the character strings written on paper And its confirmation.

First Embodiment

FIG. 3 shows the electrical configuration of the character recognition device.

The CPU 10 converts the input image data captured by the camera 1 into two
It performs digitization, scanning of an extraction window, calculation of a feature amount, determination of a cutout position, and character recognition processing of the cutout image.

The I / O control device 11 captures the image input by the camera 1 and stores the image in the image storage RAM 12
Input control transmitted via line 19 and CPU 1
For example, output control is performed to output the result of the recognition process using a 0 to a display device.

The image storage RAM 12 has a bus line 1
9 is a memory for temporarily storing input image data transmitted from the I / O control device 11 through the CPU 9 and binary image data subjected to the binarization processing in the CPU 10.

The calculation RAM 13 is a memory for temporarily storing a feature obtained by scanning the extraction window and a feature used for character recognition processing.

The binarization program ROM 14 is a read-only memory that stores a program necessary for binarizing an input image.

The extraction control program ROM 15 scans an extraction window stored in an extraction window ROM 16 to be described later with respect to the binarized binary image data, calculates a characteristic amount, and determines a character extraction position. Is a read-only memory for storing a program necessary for performing the above operation.

The extraction window ROM 16 is a read-only memory that stores an extraction window corresponding to a character arrangement of a target character string assumed in advance.

The character recognition program ROM 17 stores a calculation RA of the input image stored in the image storage RAM 12.
It is a read-only memory that stores a portion corresponding to the character extraction result stored in M13 and a program necessary for the character recognition process performed by the CPU 10 based on the reference pattern stored in the character recognition dictionary ROM 18 in advance.

The character recognition dictionary ROM 18 is a read-only memory in which reference patterns necessary for character recognition processing are stored in advance.

The above I / O control device 11 and image storage R
AM12, arithmetic RAM13, binarized program ROM
14, a cutout control program ROM 15, an extraction window ROM 16, a character recognition program ROM 17, and a character recognition dictionary ROM 18
Is connected to the CPU 10 via the.

The characteristic value calculation processing and cutout position determination processing performed by the CPU 10 in the character recognition processing will be described with reference to FIG.

FIG. 4A shows image data obtained by binarizing input image data containing a character string whose character arrangement is known.

If the character string binarization results in blurring 20 and noise 21 as shown in the figure due to input conditions or failure of binarization, the character string using labeling and projection length as in the prior art is used. With the clipping method, it is difficult to determine an accurate character clipping position, and as a result, a problem often occurs in character recognition.

Therefore, in order to extract a character string from the binary image data as shown in FIG. 4A, the feature amount is calculated by scanning the extraction window 22 shown in FIG. Determine the cutting position. The extraction window 22 shown in FIG. 4B is for extracting a four-digit character string. A hatched portion 23 is a character area, and a surrounding blank portion 2 is shown.
4 respectively correspond to the background areas. The position of the extraction window 22 to be scanned is represented by a point 25 at the upper left corner thereof (this is referred to as a scanning point).

FIG. 4C shows a state in which the above-described extraction window 22 is being scanned in order to determine the cut-out position of a 4-digit character string from the binary image data.

The feature quantity is represented by a result obtained by subtracting the number of all black pixels included in the background area 24 from the number of all black pixels included in the character area 23. The calculation of this feature is
This is performed for each of the scanning points 25. Generally, the extraction window 22 is scanned in the raster direction, that is, from the upper left of the image data, and is scanned from left to right and from top to bottom.

FIG. 4D shows a state in which the extraction window 22 is located at the position 26 where the feature value has the maximum value. By extracting characters in the character area 23 of the extraction window 22 at this position, correct character recognition processing can be performed. The position of the extraction window 22 at which the feature value has the maximum value is called a character cutout position.

FIG. 5 is a flowchart showing a character recognition process in the character recognition device.

The image of the license plate 6 of the vehicle captured by the camera 1 shown in FIG. 1 is input as image data via the I / O control device 11 (step 30) and stored in the image storage RAM 12. A binarization process is performed on the stored image data based on a program stored in the binarization process program ROM 14 (step 3).
1) The binarized image data is stored in an image storage RAM 12
Is stored.

Next, the binary image data is scanned by the extraction window 22 stored in the extraction window ROM 16 to calculate the characteristic amount for each scanning point 25. The obtained characteristic amount is stored in the arithmetic RAM 13 (Steps 32-36). Here, the variable i is a counter indicating the scanning point 25 described above, and the maximum value of the scanning point 25 is N.
That is, the scanning of the extraction window 22 starts at the upper left scanning point i = 1 of the image, scans from left to right, from top to bottom, and ends at the lower right scanning point i = N. The feature amount is calculated by moving the extraction window 22 to the scanning point indicated by the counter i (step 33) and subtracting the total number of black pixels included in the background area 24 from the total number of black pixels included in the character area 23. Are stored in the operation RAM 13 (step 34). This process is repeated until the entire binary image data has been scanned (step 35).

When the scanning is completed and the characteristic amounts at all the scanning points are calculated, the largest characteristic amount among the characteristic amounts stored in the arithmetic RAM 13 is calculated based on the program stored in the cutout control program ROM 15. Are found, and the position of the scanning point that caused the characteristic amount is determined as a character cutout position (step 37). The result is stored in the operation RAM 13, and image data included in the character area 23 of the character extraction position extraction window 22 is extracted from the input image data stored in the image storage RAM 12. That is, it is extracted for character recognition processing (step 38).

Finally, based on the program stored in the character recognition program ROM 17 and the reference pattern stored in the character recognition dictionary ROM 18, a character recognition process is performed on the extracted image data (step 3).
9). Character recognition processing is performed by normal pattern matching or other methods. Character recognition result is I
Output via the / O control device 11 (step 40),
All processing ends.

Second Embodiment

The second embodiment is obtained by adding a cutout position adjusting process to the process of the first embodiment.

The electrical configuration of FIG. 3 is used as it is in the second embodiment.

Depending on the positional relationship between the camera 1 and the license plate 6 to be recognized, the character spacing on the license plate 6 may become narrower or wider on the image. In the first embodiment, the extraction window 2
2 are predetermined in advance. Therefore, as shown in FIG. 6A, a state may occur in which all characters do not fit in all the character areas 23 even at the correct cutout position. In such a case, accurate character recognition processing cannot be performed.

Then, after the extraction window 22 is brought to the cutout position by the method of the first embodiment, the extraction window 22 is divided for each character. That is, FIG.
As shown in the figure, an extraction window 50 is provided for each character in a range including the character region 23 of the extraction window 22, and a character region 5 corresponding to the character region 23 is provided in the extraction window 50.
Set 1. In the extraction window 50, a portion outside the character region 51 is defined as a background region 52.

The feature amount is calculated for each extraction window 50. The total number of black pixels included in the background area 52 is subtracted from the total number of black pixels included in the character area 51, and the result of the subtraction becomes a feature amount in the extraction window 50. As shown in FIG. 6C, the position of each extraction window 50 is finely adjusted by moving the extraction window 50 up and down and left and right so that the feature amount is maximized.

FIG. 7 shows a character recognition process according to the second embodiment.
6 is a flowchart corresponding to FIG.

The processing in steps 30 to 37 is the same as each processing in FIG.

In step 37, the extraction window 22
Is positioned, the extraction window 50 divided for each character is used by using the extraction window 50 divided as described above.
The position is finely adjusted (step 60). When the extraction position is determined for each extraction window 50, each extraction window 5
Image data in the character area 51 in 0 is cut out (step 38), and character recognition processing is performed based on the cut out image data (step 39).

Third Embodiment

In the third embodiment, processing using a plurality of extraction windows is added to the processing of the first embodiment in order to cope with the fact that the character size of the license plate differs depending on the vehicle type.

The electrical configuration of FIG. 3 is used as it is in the third embodiment.

As shown in FIG. 8, there are two types of vehicle license plates, a license plate 6A for a large vehicle and a license plate 6B for a medium-sized vehicle. It is necessary to prepare two types according to the size of the plate.

Therefore, a plurality of (two in this case) extraction windows (2
2A and 22B) are prepared, each extracted window is scanned, and the characteristic amount is calculated at each scanning point.
The position of the extraction window in which the one showing the maximum value among all the feature values is generated is determined as the cutout position.

FIG. 9 and FIG. 10 are flowcharts showing the character recognition processing of the third embodiment and corresponding to FIG.

A counter j for scanning a plurality of types of extraction windows stored in advance in the extraction window ROM 16 is provided. The counter j is set to j = 1 (step 80), the binary image is scanned using the extraction window selected by the counter j (step 81), and the feature amount is calculated at each scanning point (step 32). ~ 36).

In this processing, while the counter j is incremented (step 83), the stored n extraction windows (n = 2 in the case of the license plate described above) are stored.
(Step 82). The cut-out position is determined to be the calculated feature amount, that is, the position of the extraction window that produces the maximum value among the feature amounts of all the scanning points N × all the extraction windows n.

Fourth Embodiment

The electrical configuration shown in FIG. 3 is used as it is in the fourth embodiment.

FIGS. 11 and 12 are flow charts showing the character recognition processing of the fourth embodiment and corresponding to FIG. Here, it is assumed that the extraction window 22 of the first embodiment is used.

The processing of steps 30 to 36 is the same as that shown in FIG.

In step 90, based on the program stored in the cut-out control program ROM 15, from among the characteristic amounts obtained by scanning the extraction window 22 stored in the arithmetic RAM 13, the upper m Are selected as cutout position candidates.

Next, for each selected candidate, the character string of the character string is referred to while referring to the image data contained in the character area at each candidate position of the input image and the binary image stored in the image storage RAM 12. The suitability as the cutout position is calculated (steps 91 to 94). A candidate position where the suitability has the maximum value is determined as a character cutout position (step 95). Compatibility is a value that represents characteriness.
Specifically, it is represented by the total number of black pixels included in the character area 23, the character feature amount, and the like.

Fifth Embodiment

The electrical configuration of FIG. 3 is used as it is in the fifth embodiment.

FIGS. 13 and 14 are flowcharts showing the character recognition processing of the fifth embodiment and corresponding to FIG.

The processing of steps 30 to 36 is the same as that shown in FIG. In addition, among the feature amounts obtained by scanning the extraction window 22, the top m
(Step 9)
0) is the same as in the fourth embodiment.

Here, priorities are determined in descending order of the calculated feature values (step 100).
According to the determined priorities, similarly to the fourth embodiment, the input image stored in the image
The suitability as the cutout position of the character string is calculated with reference to the image data included in the character area at each candidate position in the binary image (steps 91 to 94). Here, since the priority order of the candidate positions has already been determined, the suitability is compared with a predetermined threshold value (step 101). It is determined as a cutout position (step 1
02).

Sixth Embodiment

The sixth embodiment is obtained by adding a process of scanning while switching the extraction windows having different sizes depending on the scanning position to the process of the first embodiment.

The electrical configuration shown in FIG. 3 is used as it is in the fifth embodiment.

As shown in FIG. 1, when taking an image while looking down from the front in the traveling direction of the vehicle, if the viewing angle θ of the camera 1 is set relatively narrow as shown in FIG. There is almost no difference in size depending on the location depending on the distance from the object to the recognition target, and the image of the license plate 6 is not distorted. However, when the viewing angle θ is set to a large value as shown in FIG. 15A, when the vehicle is in front (in FIG. 15B, when the vehicle is on the lower side), 5A and when the vehicle is At a certain time (when the vehicle is on the upper side in FIG. 15B), the size of the target character strings (6C and 6D) is different from that of 5B, and correct character recognition processing cannot be performed.

Therefore, when scanning the extraction window,
Scanning is performed while switching a plurality of extraction windows (22C and 22D) having different sizes depending on the position on the screen, and a character cutout position is determined from the obtained feature amount.

FIG. 16 is a flow chart showing the character recognition processing of the sixth embodiment and corresponding to FIG.

In step 120, based on the program stored in the cut-out control program ROM 15, the scanning position, that is, the position corresponding to the scanning point i, is extracted from a plurality of extraction windows of different sizes stored in the arithmetic RAM 13. An extraction window having a suitable size is selected, scanning is performed while switching, and a character cutout position is determined from the obtained feature amounts.

Seventh Embodiment

The seventh embodiment deals with a white license plate and a green license plate.

FIG. 17 is a diagram corresponding to FIG. 3 and showing the electrical configuration of the character recognition device according to the seventh embodiment. The same components as those shown in FIG. 3 are denoted by the same reference numerals, and redundant description will be avoided.

The character area limited program ROM 130 is
This is a read-only memory that stores a program necessary for processing for limiting a character area from input image data stored in the image storage RAM 12.

The inverted character determination program ROM 131 stores
In the binary image data stored in the image storage RAM 12, it is determined whether or not an image included in a predetermined character area is a reverse character. It is a read-only memory that stores a program necessary for inverting the binary image data of the corresponding part stored in the RAM 12 for use.

The above character area limited program ROM 13
The 0 and inverted character determination program ROM 131 are connected to the CPU 10 via the bus line 19, respectively.

In the seventh embodiment, a process for limiting the character area and a process for inverting the binary image included in the determined character area are added to the processing of the first embodiment.

FIG. 18 shows the license plate of the vehicle.
As shown in (A), a green character string (normal character) on a white background
Is written on a white plate 6E, and a green plate 6F on which a white character string (reversed characters) is written
There are two types. Therefore, in order to perform a correct character recognition process, it is necessary to further reverse the inverted character.

FIG. 18B shows an example of input image data including a character string whose character arrangement is known but which is not known as a normal character or an inverted character. Here, first, an approximate character area (an area where characters are written) is limited, and binarization processing is performed within the limited area. It is determined whether or not the character is an inverted character based on the binarized image data. If the character is an inverted character, character recognition processing is performed after the character is inverted.

The character area is determined by calculating an edge component for each pixel or an appropriate small area of the image data input through the I / O control device 11 (calculation of edge data), and determining an edge data having a high value. This is performed by defining a range in which data (having a large edge likeness) is gathered as a character area. By extracting the edge data, information indicating the presence of a character is output irrespective of the black-and-white (green) inverted character, so that the character area can be limited.

FIGS. 19 and 20 are flowcharts showing the character recognition processing of the seventh embodiment and corresponding to FIG.

After inputting the image data (step 30), edge data is extracted from the input image stored in the image storage RAM 12 based on the program stored in the character area limited program ROM 130 (step 150). Based on this, a character area is extracted (step 151).

In step 152, based on the program stored in the binarization program ROM 14, the image data of the limited character area stored in the image storage RAM 12 is binarized and stored in the image storage RAM 12. .

In step 153, based on the program stored in the inverted character determination program ROM 130, the number of black pixels and white pixels included in the limited character area binary image stored in the image storage RAM 12 is determined. By counting, it is determined whether the character is a normal character or a reverse character. If the number of white pixels is larger than the number of black pixels, it is determined that the character is a normal character. Conversely, if the number of black pixels is larger than the number of white pixels, the character is determined to be a reversed character. If the character is an inverted character, the binary image stored in the image storage RAM 12 is inverted between black and white (step 154). In the case of normal characters, the image data is stored as it is.

In the above processing, the extraction position determination and the character recognition processing according to steps 32 to 40 shown in the first embodiment are performed on the binarized image data in the character area.

Eighth Embodiment

In the eighth embodiment, a one-dimensional extraction window is scanned in the cut area to determine a cut position.

FIG. 21 is a diagram corresponding to FIG. 3 and showing the electrical configuration of the character recognition device according to the eighth embodiment. The same components as those shown in FIG. 3 are denoted by the same reference numerals, and redundant description will be avoided.

The character line segmentation program ROM 160
For an image stored in the image storage RAM 12,
This is a read-only program that stores a program necessary for a process of extracting a character line.

The extraction control program ROM 15A is a read-only program that stores a program necessary for performing extraction processing using the one-dimensional extraction window.

The extraction window ROM 16A is a read-only program that stores a one-dimensional extraction window.

The above character line cutout program ROM 16
0, the extraction control program ROM 15A, and the extraction window ROM 16A are connected to the CPU 10 via the bus line 19, respectively.

The character line cutout processing according to the eighth embodiment will be described with reference to FIGS. 22, 23 and 24.

The horizontal projection length data 141 is calculated for the binarized image data (step 180). A character line is cut out based on the horizontal projection length data 141 (step 181). Horizontal projection length data 1
A range V1 where 41 is equal to or larger than a predetermined value is cut out.

The vertical projection length data 140 is calculated within a range V1 determined by character line segmentation.

The one-dimensional extraction window 170 is used. This one-dimensional extraction window 170 has a character range 171.
And a background range 172 therebetween. The one-dimensional extraction window 170 is scanned in the horizontal direction in a predetermined range.

At each scanning position, the total sum of the vertical projection length data belonging to the background range 172 is subtracted from the total sum of the vertical projection length data belonging to the character range 171 in the one-dimensional extraction window 170 to obtain a feature amount (step 18
2, 34).

The position of the one-dimensional extraction window 170 at which the feature value has the maximum value is the range V2 from which the image is finally cut out (steps 37 and 38).

A character recognition process is performed based on the image data belonging to the area extracted by the ranges V1 and V2 (steps 39 and 40).

In this embodiment, the one-dimensional extraction window 1
Since the scan 70 only needs to be performed once in the horizontal direction, the cutout processing is quick.

Ninth Embodiment

In the ninth embodiment, similar to the eighth embodiment, the one-dimensional extraction window is scanned in the cut-out range to determine the cut-out position, but the processing range is determined before the binarization process. , The amount of memory used for the binarization processing can be saved.

The character line cutout processing according to the ninth embodiment will be described with reference to FIGS. 25, 26 and 27.

First, image data as shown in FIG. 25A is input (step 30). Here, instead of performing binarization processing on the entire input image data, a character line is cut out and binarization processing is performed only on the character line.

In step 200, the image storage RA
With respect to the input image data stored in M12, horizontal edge data is extracted by a program stored in the character line extraction program 160, and character extraction is performed based on the data.

In step 31, binarization processing is performed only on the extracted character lines, and vertical projection length data is created for the binarized image (step 20).
1), stored in the arithmetic RAM 13; FIG. 25 (B)
Indicates that the character area has been cut out and the vertical projection length data has been created.

In the same manner as in the above-described eighth embodiment, the extraction window ROM 1
One-dimensional extraction window 1 stored in 6A
70 is scanned (step 182), and the feature amount is calculated (step 34).

Tenth embodiment

The tenth embodiment shows a character recognition device more suitable for the present invention, and is a combination of a plurality of processes selected from the first to ninth embodiments.
This will be described with reference to the flowcharts in FIGS.

The image of the license plate 6 of the vehicle taken by the camera 1 shown in FIG. 1 is input as image data via the I / O control device 11 (step 30).

Edge data is extracted from the input image data (step 150), and a character area is limited based on the edge data and a character area limitation program (step 151).

The image data of the portion corresponding to the character area of the input image is subjected to the binarization processing based on the binarization processing program (step 152).

Based on the inverted character determination program, the number of black pixels and white pixels included in the binary image of the limited character area is counted, and it is determined whether the character is a normal character or an inverted character (step 153). In the case of inverted characters, the stored binary image is inverted between black and white (step 154). In the case of ordinary characters, the image data is stored as it is.

In accordance with the clipping control program, the stored binary image is projected in the vertical direction to create vertical projection length data.

A counter j (the maximum value is n) for scanning a plurality of types of extraction windows stored in advance.
Here, n = 2) is set to j = 1 (step 8)
0), extraction window j is selected (step 8)
1). The counter i indicating the scanning point (the maximum value is N) is i = 1
And the extraction window having a size suitable for the position corresponding to the scanning point i is scanned, and the feature amount is calculated for each scanning point (steps 32 to 36).

From the feature quantities obtained by scanning the extraction window, the top m pieces are selected as cut-out position candidates in descending order of their values (step 90), and priority is given to the calculated feature quantity values in descending order. Determine the ranking (Step 10
0).

In accordance with the determined priority, reference is made to the image data contained in the character area at each candidate position among the stored input image and binary image, and the extraction window is finely adjusted vertically, horizontally, and Compatibility as a character string cutout position is calculated (steps 91 to 94). The suitability is compared with a predetermined threshold value (step 101). If the candidate position indicates suitability equal to or higher than the threshold value, it is determined as a cutout position of each character (step 1).
02).

From the stored input image data, image data included in the character area of the character extraction position extraction window is extracted, and based on the character recognition program and the reference pattern stored in the character recognition dictionary ROM 18.
A character recognition process is performed on the extracted image data (step 39). Character recognition result is I / O control device 11
(Step 40), and the entire process ends.

As in the tenth embodiment, it is possible to select two or more arbitrary ones from the first to ninth embodiments and appropriately combine them within a contradictory range.

[Brief description of the drawings]

FIG. 1 is a layout diagram of an embodiment in which a character recognition device is used to read a license plate of a vehicle running on a road.

FIG. 2 is a diagram showing the layout of the embodiment shown in FIG. 1 as viewed from the side of a road.

FIG. 3 is a block diagram illustrating an electrical configuration of the character recognition device according to the first embodiment.

4A is an example of an input image, FIG. 4B is an example of an extraction window, FIG. 4C is a state in which the extraction window is scanned on a binary image, and FIG. Each of the extraction windows is shown.

FIG. 5 is a flowchart illustrating a character recognition process according to the first embodiment.

FIG. 6A shows a state in which the character is shifted from the character area of the extraction window positioned at the extraction position, FIG. 6B shows a state in which the extraction window is divided for each character, and FIG. Each state of adjustment is shown.

FIG. 7 shows a character recognition process in the second embodiment,
5 is a flowchart corresponding to FIG.

FIG. 8 is a diagram comparing a license plate for a large vehicle with a license plate for a medium-sized vehicle.

FIG. 9 shows a character recognition process in the third embodiment,
Is a part of the flowchart corresponding to.

FIG. 10 is a part of a flowchart corresponding to FIG. 5, showing a character recognition process in the third embodiment.

FIG. 11 is a part of a flowchart corresponding to FIG. 5, illustrating a character recognition process in the fourth embodiment.

FIG. 12 is a part of a flowchart corresponding to FIG. 5, illustrating a character recognition process in the fourth embodiment.

FIG. 13 is a part of a flowchart corresponding to FIG. 5, showing a character recognition process in the fifth embodiment.

FIG. 14 is a part of a flowchart corresponding to FIG. 5, illustrating a character recognition process in the fifth embodiment.

15 (A) is a view of the arrangement shown in FIG. 1 as viewed from the side of the road 4, in a state where the visual sense of the camera is set to be larger than in FIG. 2, and FIG. 15 (B) is in FIG. It is the figure seen from the position of the camera shown, and shows the state from which the size differs according to the distance depending on the distance.

FIG. 16 is a flowchart corresponding to FIG. 5, illustrating a character recognition process in the sixth embodiment.

FIG. 17 is a block diagram illustrating an electrical configuration of a character recognition device according to a seventh embodiment.

18A is a comparison between a license plate of a normal character (green on a white background) and a license plate of a reversed character (white on a green background), FIG. 18B is an input image, and FIG. 18C is a character region cut out. Indicates the status.

FIG. 19 is a part of a flowchart corresponding to FIG. 5, illustrating a character recognition process in the seventh embodiment.

FIG. 20 is a part of a flowchart corresponding to FIG. 5, illustrating a character recognition process in the seventh embodiment.

FIG. 21 is a block diagram illustrating an electrical configuration of a character recognition device according to an eighth embodiment and a ninth embodiment.

FIG. 22 shows a state in which a character line is cut out from a binary image using horizontal projection length data and vertical projection length data is extracted in the region, and one-dimensional extraction for scanning over vertical projection length data. Here is an example of a window.

FIG. 23 is a part of a flowchart corresponding to FIG. 5, illustrating the character recognition processing in the eighth embodiment.

FIG. 24 is a part of a flowchart corresponding to FIG. 5, illustrating a character recognition process in the eighth embodiment.

FIG. 25A shows a state in which a character line is cut out from an input image using edge data, and FIG. 25B shows a state in which vertical projection length data is extracted in that area; 5 shows an example of a one-dimensional extraction window for scanning over data.

FIG. 26 is a part of a flowchart corresponding to FIG. 5, illustrating the character recognition processing in the ninth embodiment.

FIG. 27 is a part of a flowchart corresponding to FIG. 5, illustrating the character recognition processing in the ninth embodiment.

FIG. 28 shows a character recognition process in a tenth embodiment.
It is a part of the flowchart corresponding to FIG.

FIG. 29 shows a character recognition process in a tenth embodiment.
It is a part of the flowchart corresponding to FIG.

FIG. 30 shows a character recognition process in a tenth embodiment.
It is a part of the flowchart corresponding to FIG.

FIG. 31 shows a character recognition process in a tenth embodiment.
It is a part of the flowchart corresponding to FIG.

[Description of Signs] 1 TV (video) camera 5 Vehicle 6 License plate 22 Extraction mask 23 Character area 24 Background area 25 Scanning point 170 One-dimensional extraction window

Continuation of the front page (56) References JP-A-3-97085 (JP, A) JP-A-63-136179 (JP, A) JP-A-1-213769 (JP, A) JP-A-3-265985 (JP) JP-A-63-318688 (JP, A) JP-A-5-28317 (JP, A) JP-A-63-257883 (JP, A) JP-A-63-158677 (JP, A) (58) Field surveyed (Int. Cl. ⁷ , DB name) G06K 9/34 G06K 9/20 340

Claims (2)

(57) [Claims] An input unit for inputting an image including a target character string whose character sequence is known; a binarizing unit for binarizing the input image; an image data binarized by the binarizing unit; On the other hand, a feature amount calculation for scanning a plurality of types of extraction windows of different sizes each having an arrangement corresponding to a known character arrangement and calculating a feature amount for each position of the extraction window for each of the plurality of types of extraction windows Means, a cut-out position determining means for determining a character cut-out position based on the feature amount obtained by the feature amount calculating means, and image data cut out according to the position extraction window determined by the cut-out position determining means, Character recognition means for performing character recognition processing based on a pre-stored reference pattern, and Output means for outputting a result, the character recognition apparatus having a. 2. An input unit for inputting an image including a target character string whose character sequence is known, a binarizing unit for binarizing the input image, and an image data binarized by the binarizing unit. On the other hand, a plurality of extraction windows of different sizes having a layout corresponding to a known character arrangement are scanned by switching according to the position on the image, and a feature amount is calculated for each position of the extraction window. Calculating means, a cutout position determining means for determining a character cutout position based on the feature amount obtained by the feature amount calculating means, and image data cut out according to the position extraction window determined by the cutout position determining means. Character recognition means for performing character recognition processing based on a reference pattern stored in advance, and an output means for outputting a result obtained by the character recognition means. , The character recognition device equipped with.