JPS596419B2 - Character extraction method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a character extraction method for a character recognition device, and more particularly to a character extraction method that can effectively remove noise other than characters.

Generally, characters printed in one horizontal line are scanned in the vertical direction by a sensor having photoelectric conversion elements arranged in a vertical row, and the sensor or characters are moved horizontally to sequentially read the characters.

In Figure 1, the field of view 2 of the sensor is shown on paper 1 on which characters are printed.
Indicates the situation of moving above. Paper 1 has the numbers ``1'', ``2'', and ``3'' printed on it as characters 3, 4, and 5 to be read, as well as stains 6, T, and 8 around it. These stains are other than character information to be recognized, and become noise, causing misreading or unreadability. Conventionally, such noise has been removed by changing the reading level, installing a special filter in front of the sensor, or by reading the noise once and then processing it using software to remove it. was.

However, the former method has unreliable operation, and the latter method requires a large amount of program capacity and processing time, and even then, effective noise removal cannot be achieved. SUMMARY OF THE INVENTION An object of the present invention is to eliminate the above-mentioned drawbacks of the prior art and to provide a cutting method that is highly effective in removing noise.

Another object of the present invention is to provide a method for cutting out characters so that they can be reliably distinguished from adjacent characters.

Still another object of the present invention is to reduce the burden on the computer of the recognition unit by realizing extraction using hardware.
This is a cutting method that improves reading speed. Still another object of the present invention is to provide a method for cutting out images in such a way that even if the scanning speed differs, the scanning speed is not affected much. The present invention calculates the appearance frequency distribution of character line segments projected onto the vertical axis of the read pattern every time one scan in the vertical direction is completed,
This method is characterized by normalizing the frequency distribution, finding vertically continuous line segments of a certain length from the normalized data, and cutting out the height of the character from the length range of the line segments.

The present invention will be explained below with reference to the drawings. FIG. 2 shows, as an embodiment of the present invention, a character recognition apparatus in which a scanner 12 is held in the hand and moved in the horizontal direction to perform manual scanning.

A sheet of paper 11 is illuminated with lamps 13a and 13b, and a character pattern on the sheet of paper 11 is imaged through a lens system 14 onto a sensor 15 having photoelectric conversion elements arranged in a row.

Each photoelectric conversion element of the sensor 15 will be referred to as a cell, and the area where data is compressed to the required resolution will be referred to as a unit area. Since the light reflected from the background area of the paper 11 and the light reflected from the character area are different, the corresponding signal levels from each cell are different. The signals obtained from each cell of the sensor 15 are applied to a control and binarization circuit 16, and black and white are determined by level determination. For example, the signal corresponding to the background area of paper 11, that is, white, is
The counter control circuit 25 receives signals corresponding to each unit area sequentially from the control and value conversion circuit 16. When the signal is black, that is, "17", the output of the shift register 27 is preset to the preset counter 26, and the content of the preset counter 26 is incremented by 1, and the result is stored in the shift register 27.

Also, when the signal is white, that is, "0", the preset counter 26
The output of the shift register 2r is preset to , and the contents of the preset counter 26 are directly transferred to the shift register 27
It is controlled so that the data is stored in the . Further, when the one-character scanning end signal GAP, which will be described later, is input from the gear flag 21, the counter control circuit 25 clears the contents of the preset counter 26, and when the signal is 7, adds 1 to "O". When , the preset counter 2 is used as is.
Control is performed so that the contents of 6 are stored in the shift register 27.

The shift register 27 registers the number of stages corresponding to the number of cells of the sensor 15 or the number of unit areas corresponding to the data compressed in the vertical direction. The number of times black appears in each area is counted, and the contents of the shift register 27 represent the number of times a horizontal fork line segment appears in each unit area. Further, when the contents of the preset counter 26 are stored in the shift register 27, the contents of the preset counter 26 are sent to the comparison circuit 28. The comparison circuit 28 compares the contents of the storage device 29 (assumed as A) with the preset counter 26.
The contents inputted from the input terminals (referred to as B) are compared, and B is stored in the storage device 29 only when B>A. When one vertical scan is completed, the signal SEN from the control and binarization circuit 16
At timing D, the contents of the storage device 29 are transferred to the maximum value storage device 30, and then the contents of the storage device 29 are cleared.

That is, the maximum value storage device 30 stores the maximum value of the counted number of black appearances for each unit area stored in the shift register 27.

The normalization circuit 31 calculates the ratio between the contents of the shift register 2r and the contents of the maximum value storage device 30, and divides it into a plurality of ratios. (Hereinafter referred to as normalized data) When moving by hand, the scanning speed varies greatly depending on the individual or the same person, but in the present invention, the scanning speed is determined based on the maximum number of times black appears in each unit area. Since the number of appearances of black is normalized, the normalized value is not affected greatly even if the horizontal scanning speed differs. The normalized and divided contents are sent to the binarization circuit 32 or the identification circuit 35.

The sorting circuit 35 identifies characters from the normalized data, for example, based on the arrangement of the normalized data. Also: value conversion circuit 32
Then, the normalized data is compared with a certain threshold value SHL, and the binary information, that is, the pattern information projected on the vertical axis is divided into a character area and a background area of the paper 11.

For example, in FIG. 1, the sensor field of view 2 is from the position t1 to T
FIG. 5 shows the normalized data at the T2 position when moving to the T2 position, and FIG. 6 shows the binarized normalized data.

The normalized data corresponding to character 3 and stains 4 and 5 is 61
and 60,62.

In FIG. 6 after binarization, the background area and character area are represented by W (white) and B (black), respectively. The read-only storage device 33 is composed of the output signal of the binarization circuit 32 and the output contents of the storage device 34, and the read contents are written in the storage device 34, and the signals SET, . Output POUT.

The memory device 34 is cleared by the 1 vertical scan end signal SEND.

FIG. 4 shows a state transition diagram of the contents read from the read-only storage device 33.

Assume that a signal as shown in FIG. 6 is input from the binarization circuit 32. SO and Sn indicate unit areas at the uppermost and lowermost ends of the sensor field of view 2. The binary signal from the upper end of the unit area is B, W, W, W, W, W, B, B, B, B, B, B, W
, W, W, B, W, W, W, W, W, , W, W, W, the states are 0B, 1 in FIG.
B, 1W, 0B, 1B, 2B, 3B, 4B, 5B, 6B
, 6W, 0B, 1B, 0B. Also SET. ．． P.O.
UT represents the signal that is output when states 1B, 5W, and 6W are reached, the signal SET means the start of cutting out characters, and the signal POUT means the end of cutting out characters, and the signal SET is output many times. When a signal is sent, priority is given to the signal sent. In the binarized data shown in Figure 6, 1B is 1
The signal SET is output, the signal SET is output again, and then POUT is output.
is output, indicating that vertical character cutting has been completed. The control and binarization circuit 16 has a counter to determine which unit area is currently selected during vertical scanning, and the content SCC of this counter indicates the address SCC of the unit area.

In FIG. 3, a storage device 49 is a read-only storage device 33.
When the signal SET is input from the cell, the cell address SCC at that time is written. Since priority is given to the signal output later, in the binary data of FIG. 6, the address SCC at that time is written again at the second SET signal. In addition, the contents of the storage device 34, that is, the current state Ci (1B
, 2B, etc.) to the storage device 51.
If there is noise of the same level as the characters, the extraction is completed once and then the extraction is performed again in the same manner.

For example, when the sensor field of view 2 moves to position T3 in Figure 1, the normalized data of character 5 and noise 8 are 63 and 63 in Figure 7, respectively.
It becomes 64. The extraction is finished using the normalized data 63,
SCC at that time. Ci is written to the storage devices 49 and 51, and when the extraction is finished using the normalized data 64, the contents of the storage devices 49 and 51 are written to the storage device 50, respectively.
, 52, the next SCC, Ci is transferred to the storage device 49,
Write to 51.

In FIG. 3, the explanation has been made with two cuts, but the same holds true for two or more cuts. The selection circuit 53 selects a character area from among a plurality of extracted character areas, and stores the contents of the storage device 38, that is, the address SCC when it was extracted in the previous vertical scan, and the storage device 51, 52, that is, the character area is determined from the state Ci at the time of cutting out, and the character area is determined from the storage device 49 or 5.
0 is written to the storage device 54.

For example, the character area is determined based on the fact that there is no large difference between the scanned character area and the adjacent character area, and on the basis of the condition Ci that is closest to the standard height of the character.

If the POUT signal has been output at least once when one vertical scan is completed, the selection circuit 53 continues to output the CK2 signal, and transfers the contents of the storage device 54 to the storage device 38 at the timing of CK1 immediately before the arrival of the SEND signal. By writing, a character is cut out in the vertical direction and the address CTA of the unit area at the uppermost end of the character is stored in the storage device 38.

The signal SET is logically ANDed with the signal CK output for each scan of the unit area by the AND gate 39, and is input to the storage device 49. When one vertical scan is completed, the contents of the storage devices 49, 50, 51, and 52 are cleared.

Further, when the output of the AND circuit 40 becomes logic "1", a cutout flag 42 is set to indicate that a character is cut out in the vertical direction.

When the signal ER, which is caused by noise during cutting, or the space signal SP, which will be described later, is output, the OR gate 4
5 clears the cutout flag 42 and assumes that no character is cut out. If the characters have not been cut out, the characters are not identified to avoid erroneous recognition.
When detecting the end of scanning for one character, the arithmetic circuit 44 adds a constant (FIG. 6 d) to the address of the cell and compares it with a margin above the vertically cut out range of the character. Output to circuit 41.

The comparison circuit 41 compares the unit area address CTA of the uppermost end of the cut out character with the content sent from the arithmetic circuit 44, and when both sides are equal (the upper end cell is set to O, and the lower cells have addresses) is large), character height detection flag 4
Outputs the signal to 3. The character height detection flag 43 is set when the cutout flag 42 is set and a signal is input from the comparison circuit 41, and outputs a signal POF. The counter 47 counts the clock signal CK after the signal POF becomes logic 21'' until a unit area corresponding to the character height CHH (see FIG. 6) has been scanned, and when it has counted CHH, it outputs an overflow signal. The character height detection flag 43 to be output is an OR signal (OR circuit 48) of this overflow signal and the 1-column vertical scanning end signal SEND.
) is cleared by For example, we know that the content of the SCC during vertical scanning is Sn, and the height of the character is h.)
In FIG. 6, when CHH=h+2d CTA+h+d≧Sn≧CTA-d, the character is in operation, and the signal POF is output.

Counter 4r is also cleared by signal SEND.

With the above operations, the character position in the vertical direction is cut out as shown as POF in FIG. 6. Next, we will discuss cutting out in the horizontal direction. Gap detection flags 18 and 19 are set by the SEND signal generated after vertical scanning, and gap up detection flag 18 is cleared when black (character line segment) is detected between the vertical cutting signals POF. , a gap detection flag 19 is a flip-flop that is cleared when black (character line segment) is detected within one vertical scan. Immediately before generating the SEND signal after one vertical scan, the gap detection flag is set to 1 at the timing of clock CKl.
Take the logical sum of the output signals of 8 and 19 (logical sum circuit 20)
This logic signal is stored in the gap flag 21 and the contents of the gap flag 21 are transferred to the gap flag 22. In other words, the gap flag 21 is a flag that is set when no character line segment is detected in the vertical character extraction section (POF) or when no character line segment is detected at all within one vertical scan.
is used to store the contents of the gap flag 21 during the previous vertical scan. Output signal GAP of gap flag 21
is output to the counter control circuit 25, and is ANDed with the signal SEND by the AND gate 23, and the contents of the space counter 24 are added. The space counter 2
When the content of 4 exceeds a preset constant, it is considered that one line of characters has been scanned and a space signal SP is output.
Initial state. The space counter 24 is cleared when black (character line segment) is detected. The gap flag 22 outputs the signal BFGAP and stores the signal GAP.

At this time, when BFGAP is "1" and GAP is 202, it is assumed that scanning of one character has been completed, and the character is identified. In FIG. 1, when the sensor field of view 2 moves to position T2, if it is cut out as shown in FIG. 6, the gap detection flag 19 is /'0'', but the gap detection flag 18 is ``1''. Then, the gap flag 21 is set to 7"5.

If the gap flag 21 is set to "O" at the position before T2, then at T2 GAP="17BFGAP='207 and 1
It assumes that character scanning is completed and then performs identification operations.

The space counter 24 determines that one line ends when the number of vertical scans is much greater than the number of vertical scans between characters. As explained above, in the present invention, the appearance frequency distribution of character line segments is determined for each unit area projected onto the vertical axis for each vertical scan, the number of appearances for each unit area is normalized by the maximum number of appearances, and then the character The feature is that it is not affected by the scanning speed because it cuts out images.

Furthermore, by normalizing and appropriately selecting the threshold value SHL8, value conversion is performed, so that noise around characters can also be removed.

In the description, the shift register 27 has the same number of stages as the number of cells of the sensor 15, but it is also possible to reduce the number of stages of the shift register 27 by compressing data or the like.

In addition, the fourth factor is determined based on the characters to be read and the dimensions of the sensor.
The state transitions shown in the diagram are different.

The above explanation deals with the case where the scanning speed in the horizontal direction is variable, but for example, when scanning is performed mechanically at a constant speed, the count value of the number of appearance of character line segments for each unit area and the maximum value of the number of appearance The same is true if you do not calculate the ratio of , but only the count value.

[Brief explanation of the drawing]

Figure 1 is a diagram of the relationship between read characters and read heads, Figures 2 and 3 are block diagrams of the present invention, Figure 4 is an explanatory diagram of a read-only storage device, and Figures 5 and 7 are character line segments. The appearance frequency distribution diagram, FIG. 6 shows the binarization change diagram. Explanation of symbols, 1...Paper, 2...Sensor, 3...Character, 4,5...Noise,
11...paper, 12...scanner, 1
3a, b... Lamp, 14... Lens system, 15... Sensor, 16... Control,
and binarization circuit, 17, 23, 39, 40, 46...
...AND gate, 20,45,48...OR gate, 18,19...Gap detection flag, 21,22...Gap flag, 24...
... Space counter, 25 ... Counter control circuit, 26 ... Preset counter, 27.
... Shift register, 28 ... Comparison circuit, 29 ... Storage device, 30 ... Maximum value storage device, 31 ... Normalization circuit , 32...
...: Value conversion circuit, 33... Read-only storage device,
34...Storage device, 35...Identification circuit, 3r, 38...Storage device, 41...
Comparison circuit, 42... Cutting flag, 43...
... Character height detection flag, 44 ... Arithmetic circuit, 47 ... Character height counter, 50, 51, 5
2, 54... Storage device, 53... Selection circuit.

Claims (1)

[Claims] 1. In a character recognition device that scans characters, symbols, etc. with a sensor having photoelectric conversion elements arranged vertically or horizontally, and identifies the scanned characters, symbols, etc., character line segments of the sensor are used. The number of occurrences of is counted for each unit area, and the appearance frequency distribution of character line segments for each unit area in the sensor array direction is calculated.
The character area in the direction of the sensor array is detected based on this appearance frequency distribution, and if multiple character areas are detected in the direction of the sensor array, one is selected from them to detect the character area in the direction of the sensor array. A character cutting method characterized by cutting out an area. 2. Characters, symbols, etc. that are scanned by a sensor having photoelectric conversion elements arranged vertically or horizontally;
In a character recognition device that identifies symbols, etc., the number of occurrences of character line segments on the sensor is counted for each unit area, the appearance frequency distribution of character line segments for each unit area in the sensor array direction is calculated, and the frequency distribution of character line segments for each unit area is calculated. Detect the maximum number of occurrences of character line segments, find the appearance frequency distribution normalized by the ratio of the maximum value to the number of occurrences counted for each unit area, and detect multiple numbers in the sensor row direction. 1. A character extraction method characterized in that when a character area is detected, the character area is extracted by selecting one of the character areas.