JPH0816919B2 - Concavo-convex pattern reader - Google Patents

Description

The present invention relates to a concavo-convex pattern reading device for reading a concavo-convex pattern formed on, for example, a plastic card, a metal plate, or the like.

[Prior Art] FIG. 2 is an explanatory view showing a conventional concave-convex pattern reading device disclosed in Japanese Patent Application Laid-Open No. 60-144884.

In the figure, 1 is a plastic card as an information carrier, and characters, figures, etc. are engraved in the form of concave portions or convex portions on the plane portion thereof. Here, the case where the letter L is stamped is shown.

Reference numerals 2a to 2d are lamps that sequentially irradiate the card 1 with light from four different oblique directions. A television camera 3 captures an image of the card 1 illuminated by the lamps 2a to 2d. An A / D conversion circuit 4 digitizes the output signal of the television camera 3. Reference numeral 5 denotes a switch, which is A / when illuminated by the lamps 2a to 2d.
The output of the D conversion circuit 4 is sequentially switched to each of the memories 6a to 6d and output. Reference numeral 7 is an OR gate, which outputs a logical sum signal of the outputs of the memories 6a to 6d to the memory 8.

In the above structure, the unevenness of the card 1 illuminated by the lamp 2a is photographed by the television camera 3, this image signal is digitized, and the memory 6a is formed via the switch 5.
Enter and memorize. The image stored at this time is the shaded portion that the light of the lamp 2a does not reach.

Similarly, the image when illuminated by the lamp 2b is stored in the memory 6b and the image when illuminated by the lamp 2c is stored in the memory 6b.
The image when illuminated by the lamp c and the lamp 2d is stored in the memory 6d. These images are the shadow areas formed by the corresponding lamps.

The outputs of the memories 6a to 6d are added via the OR gate 7 and input to the memory 8. Therefore, the memory 8 stores the total sum of the shadow portions formed by the lamps.

[Problems to be Solved by the Invention] However, the above conventional example has the following problems.

That is, when the groove width of the embossed character of the card 1 is wide, the shadow due to the illumination does not spread over the entire groove width, so that only the binary image of the peripheral portion of the groove can be obtained, and the character is accurately read. There was a problem I couldn't do.

In addition, depending on the position of the character on the card 1, the lamp 2a to
There was a problem that accurate reading could not be done because the illumination conditions by 2d were different.

Further, there is a problem in that if the background color of the surface to be read of the card 1 is a dark color, or if there is a pattern, printed characters, etc., there is a risk of misreading.

Furthermore, in the above-mentioned conventional example, since it is necessary to illuminate the card 1 from four different directions, four lamps 2a to 2d are required, which causes a problem that the device becomes large.

Therefore, the present invention has been made in order to solve the problems of the conventional techniques as described above.
An object of the present invention is to provide a concave / convex pattern reading device capable of performing correct pattern recognition regardless of the position and width of the concave / convex pattern and capable of downsizing.

[Means for Solving the Problem] A concave-convex pattern reading device according to the present invention irradiates an information carrier carrying information by a concave-convex pattern formed on a surface to be read with light from two different directions. And a second light source, reflected light from the information carrier when illuminated by the first light source, and reflected light from the information carrier when illuminated by the second light source. And a reading signal of the reading sensor when the information carrier is optically illuminated by the first light source, and a reading sensor when the information carrier is optically illuminated by the second light source. When the reading signal of the reading sensor is converted into a digital signal, and the digital signal values are Da and Db, and a predetermined threshold value is S, Da and Db are Article shown by Da-Db | ≦ S It is judged whether the condition 1, the condition 2 shown by Da-Db> S, or the condition 3 shown by Db-Da> S is satisfied, and based on this judgment result, a ternary signal that outputs a ternary signal The thinning processing means, the first thinning processing means for extracting the first thin line which is the boundary of the uneven pattern of the information carrier from the area satisfying the condition 2 based on the ternary signal; Second thinning processing means for extracting a second thin line which is a boundary of the concave-convex pattern of the information carrier from the region satisfying the condition 3 based on the ternary signal, and the extracted first and second thin lines. It has a connection processing means for connecting between the second thin lines.

[Operation] In the present invention, the uneven pattern of the information carrier is irradiated with light in two different directions or separately, and optical reading is performed twice at the same position of the uneven pattern. The read signal at this time is digitized, and the value of this digital signal is set to Da and Db.
And when a predetermined threshold value is S, Da and Db are shown as condition 1 represented by | Da−Db | ≦ S, condition 2 represented by Da−Db> S, and Db−Da> S. It is determined which of the following conditions 3 and 4 is satisfied, and a ternary signal is output based on the result of this determination.

Thus, in the present invention, based on the difference between the value Da of the digital signal when illuminated by the first light source and the value Db of the digital signal when illuminated by the second light source, the condition It is determined which of 1, 2, 3 is satisfied. Therefore, simply comparing the output value of the reading sensor with the reference value, the ground color (for example,
Although the black portion) was also read (that is, the shadow and the black portion could not be distinguished), according to the present invention, it is possible to detect only the shade of the uneven pattern and to detect the ground color such as the black portion. Since it is not present, it is possible to reduce the possibility of erroneously recognizing the letters or numbers represented by the concavo-convex pattern.

Here, the area satisfying the condition 1 is the flat portion (either the bottom surface of the concave portion or the upper surface of the convex portion) of the information carrier, and the area satisfying the condition 2 is when light is irradiated by the second light source. The sloped portion on the side where the shadow is formed (boundary of the concavo-convex), and the region satisfying the condition 3 is the sloped portion on the side where the shadow is formed when the light is irradiated by the first light source (the slope of the condition 2). The part has an inclined surface in the opposite direction).

Further, the first thinning processing means, based on the ternary signal,
The first thin line which is the boundary of the concave-convex pattern of the information carrier is extracted from the region satisfying the condition 2, and the second thinning processing means is based on the ternary signal, and is in the region satisfying the condition 3. A second thin line that is the boundary of the uneven pattern of the information carrier is extracted from. The connection processing means connects the extracted first and second thin lines. Thus, the first and second thin lines are extracted by the first and second thinning processing means, and then the first and second thin lines are connected by the connection processing means to obtain the connection processing. Since it is possible to bring the image information closer to the actual concavo-convex pattern of the information carrier, it is possible to reduce the possibility of erroneously recognizing the letters or numbers represented by the concavo-convex pattern. In other words, if the thin line is not extracted and the area satisfying the condition 2 and the condition 3 are satisfied.
If the region satisfying the condition is connected, the image obtained by the connecting process becomes an image with a wider outline and a blurry outline than the uneven pattern of the actual information carrier, and the letters or numbers represented by the uneven pattern are erroneously recognized. On the other hand, if the thin line is extracted and then the connection process is performed, the obtained image information can be brought close to the actual uneven pattern of the information carrier, so that the letters or numbers represented by the uneven pattern can be changed. The possibility of erroneous recognition can be reduced.

[Examples] The present invention will be described below based on illustrated examples.

FIG. 1 is a block diagram showing the configuration of an uneven pattern reading device according to the present invention.

In the figure, 10 is a sensor unit for reading the uneven pattern of the plastic card P as an information carrier. The sensor section 10 is a position symmetrical with respect to the reading position of the card P, and the line light sources 11a and 1a arranged in parallel obliquely upward with respect to the concavo-convex pattern surface of the card P.
Has 1b. Further, the sensor unit 10 includes a rod lens 12 arranged on the reading position of the card P and a rod lens 12.
It has a line sensor 13 located above 12 and a video amplifier 14.

In the sensor unit 10, only the line light source 11a is turned on to irradiate the card P with light, and the uneven pattern of the card P is read by the line sensor 14 through the rod lens 13.
Then, only the line light source 11b is turned on, and the card P
Read. The video amplifier 14 amplifies the read signal of the line sensor 13 and outputs it from the sensor unit 10. FIG. 3 shows the output of the video amplifier 14 when the line light source 11a and the line light source 11b are turned on.
FIG. 6 is an output waveform diagram showing Aa and Ab. The card P is the first
Move along the direction of the arrow in the figure to read the same
Is repeated over the entire surface.

An image processing unit 20 processes the image signal from the sensor unit 10. The image processing unit 20 converts the analog signals Aa and Ab from the sensor unit 10 into 8-bit digital signals A /
It has a D conversion unit 21 and a ternarization processing unit 22 for converting this digital signal into three values. Further, the image processing unit 20 includes 8-line memories 23a and 23b having a configuration as a shift register.
The thinning processing units 24a and 24b, the connection processing unit 25, and the one-line memory 26 for the connection processing unit 25.

FIG. 4 is an explanatory diagram for explaining the processing in the ternarization processing section 22. In the figure, the horizontal axis is the A / D converter 2
It shows the upper 8bit data Da (corresponding to the analog signal Aa) of the 16bit digital signal output from 1, and the vertical axis is the lower 8b.
The it data Db (corresponding to the analog signal Ab) is shown. In addition, the diagonal line in the upper right of the two-dot chain line in the figure is a line that establishes Da = Db, and TH1 and TH2 are lines that indicate | Da−Db | = S (S is a predetermined threshold value). .

The (0) area (hatched area) that satisfies | Da−Db | ≦ S indicates an area with a small difference in brightness when light is emitted from either the light source 11a or 11b. The flat portion of the uneven pattern is shown. The value of the threshold value S is determined by the degree of inclination considered as a flat portion.

The (+) area indicated by Da−Db> S is an area where the brightness when light is emitted from the light source 11a is brighter than the brightness when light is emitted from the light source 11b. ) Is shown.

The (−) area indicated by Da−Db <S is an area in which the brightness when illuminated by the light source 11b is brighter than the brightness when illuminated by the light source 11a, and this area is the inclined surface of the above (+) area. The inclined part which has an inclined surface facing is shown.

Therefore, the data Aa and Ab shown in FIG. 3 are plotted on the digitized data Da and Db on the diagram of FIG. 4, and the plotted points are (+) and (0 ),
Which area of (-) belongs to is determined. And
The ternary output of the (+) area, the (-) area, and the (0) area is
The data is classified into (+) area data and (-) area data and is output to the shift registers 23a and 23b, respectively.

Here, for example, if the output at the reading position is in the (+) area, the data is 1, and if it is in the (-) area, the data is 0. When the above processing is performed for 8 lines or more, (+) area data or (-) area data for 8 lines is input into the shift registers 23a and 23b. This data is input from the shift registers 23a and 23b to the thinning processing units 24a and 24b.

FIG. 5 shows thinning processing in the thinning processing units 24a and 24b, and is an explanatory diagram showing a determination pattern.

In the figure, b0 to b7 arranged side by side are shift registers 24
And the vertically arranged examples 1 to 7 show the determination patterns for the processes in the thinning processing units 24a and 24b, respectively.

The input data is sequentially shifted into the 8-bit frame of the shift registers 23a and 23b, and the oldest data is sequentially discarded. More specifically, the register 23 is used for the main scan n.
It is configured as a memory of dots × 8 lines of sub-scanning. Therefore, the data for 8 lines in the sub-scan is stored in the register 23a.
(Or 23b), the fine line position information of one line is output for the first time. After the thin line position information of one line is output, the oldest data, that is, the line data corresponding to the 7th bit in the register is discarded, and the 0th to 6th bits of data are respectively 1st to 7th.
The line data sequentially shifted to the bit position and the new ternarized line data is input to the 0th bit position.

Then, the content of the data input in the 8-bit frame of the shift registers 23a and 23b is included in any one of the determination patterns (data group of data b0 to b7 in FIG. 5) (examples 1 to 7 in FIG. 5). Judge whether there is a match. That is, the thinning processing is performed on the binary data of the (+) area data and the (−) area data of the image after the ternarization processing, and “1” is output from the thinning processing units 24a and 24b. Then, while the card P is moved and the image is sub-scanned, binary data is sequentially input to the register, and the binary data in the register is matched with any one of the judgment patterns of Examples 1 to 7 shown in FIG. For example, from the thinning processing unit 24a or 24b, "1"
Is output.

In the thin line position extraction determination pattern of the thin line processing of FIG. 5, the center of the height difference in the uneven portion of the information carrier is output as the thin line position.

The thinning process is performed on the (+) region data and the (-) region data separately, and a total of 2 bits of data is input to the concatenation processing unit 25 for each 1 bit. The thin line position data in the (+) area data is input to the Ta terminal of the connection processing section 25 and the thin line position data in the (-) area data is input to the Tb terminal of the connection processing section 25.

FIG. 6 is an explanatory diagram for explaining the contents of the thinning processing. Here, when a is the (+) data after the thinning process, b is the (−) data after the thinning process, and y is the concatenated output data, the data a is “1” and the data b is “ If it is 0 ", it means that it is a thin line that should be the beginning of the connection,
The linked output data y is set to "1".

If both the data a and the data b are "0", the thin line output data y is "1" when the thin line output data in the previous line in the one line memory (data buffer) 26 is "1".

Also, the data a is "0" and the data b is "1".
If it is, it means that it is a thin line that should be the end of the connection,
The data y is set to "0".

7 (a) to 7 (d) are explanatory views for explaining the signal processing in the present embodiment, and FIG. 7 (a) shows a card P in which the letter "B" is formed as a convex portion. (B) shows the result of reading this card P and performing ternarization processing, (c) of the figure shows the thin lines extracted by applying the thinning processing to the ternary signal, and (d) of the figure shows the thin lines. 7 shows an output pattern obtained by performing a concatenation process.

Irradiate the uneven pattern of FIG. 10A with the light sources 11a and 11b,
When the read signal at that time is digitized and the ternarization process shown in FIG. 4 is performed, as shown in FIG. 4B, the sloped portion L of the character “B” that rises to the right in FIG. Is defined as a (+) region, and the sloping portion R that is descending to the right is recognized as a (-) region. The figure (c) shows the thin line data in the (+) area and the thin line data in the (-) area, and the figure (d) is a concatenation processing output pattern.

As described above, according to the present embodiment, the uneven pattern of the card P is separately irradiated with light from two different directions, the optical reading is performed twice at the same position of the uneven pattern, and the digital signal at this time is obtained. A ternarization process is performed, the boundary line of the concave / convex pattern is extracted based on the ternary signal, and this is connected to recognize the concave / convex pattern. Since the recognition of the boundary line of such a concavo-convex pattern is not affected by the position of the character on the surface to be read, the groove width, and the background color, the concavo-convex pattern recognized by connecting the boundary lines also has these effects. There is no effect, and therefore accurate reading can be performed.

In addition, although the case where the convex portion of the concavo-convex pattern is a character is described in the above embodiment, the present invention is not limited to this, and it is also possible to read a case where the bottom surface of the concave portion is a character pattern.

[Effects of the Invention] As described above, in the present invention, the value Da of the digital signal when light is emitted from the first light source and the value Db of the digital signal when light is emitted from the second light source. Which of the conditions 1, 2 and 3 is satisfied is judged based on the difference of. Therefore, when the output value of the reading sensor is simply compared with the reference value, the ground color (for example, the black portion) is also read in the same manner as the shade of the uneven pattern (that is, the shade and the black portion cannot be distinguished. However, according to the present invention, it is possible to detect only the shadow of the uneven pattern and not the background color such as a black portion, thereby reducing the possibility of erroneously recognizing the letters or numbers represented by the uneven pattern. The effect is that you can do it.

Further, according to the present invention, the first thinning processing means is 3
Based on the value signal, the first thin line that is the boundary of the concave-convex pattern of the information carrier is extracted from the region that satisfies Condition 2,
Based on the ternary signal, the second thinning processing means extracts a second thin line which is the boundary of the uneven pattern of the information carrier from the area satisfying the condition 3, and the connection processing means extracts this. The connection between the first and second thin lines is performed. Thus, the first and second thin lines are extracted by the first and second thinning processing means, and then the first and second thin lines are connected by the connection processing means to obtain the connection processing. It is possible to make the image information closer to the concavo-convex pattern of the actual information carrier, and to obtain a blurred image with a wider contour than the actual concavo-convex pattern as in the case of connection processing without extracting thin lines. Therefore, there is an effect that it is possible to reduce the possibility of erroneously recognizing the letters or numbers represented by the uneven pattern.

 Further, the number of light sources is two, and the device can be downsized.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing an embodiment of an uneven pattern reading device according to the present invention, FIG. 2 is a configuration diagram of a conventional uneven pattern reading device, and FIG. 3 is an output waveform diagram of a video amplifier of this embodiment. FIG. 4 is an explanatory diagram of the process of the ternarization processing unit of the present embodiment, FIG. 5 is an explanatory diagram of a determination pattern for the thinning processing unit of the present embodiment, and FIG. 6 is a concatenation process of the present embodiment. 7 (a) to 7 (d) are explanatory views of the processing of the part, and the character "B" in this embodiment.
It is explanatory drawing which shows the processing content at the time of reading. 10 ... Sensor part, P ... Card, 11a and 11b ... Line light source, 12 ... Rod lens, 13 ... Line sensor, 14
...... Video amplifier, 20 ...... Image processing unit 20, 21 ...... A / D conversion unit, 22 ...... Ternary processing unit, 23a and 23b ...... Line memory, 24a and 24b ...... Thin line processing unit, 25 ...... Consolidation processing department,
26 …… 1 line memory.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Shunji Sakai 1-7-12 Toranomon, Minato-ku, Tokyo Within Oki Electric Industry Co., Ltd. (56) References JP-A-50-48856 (JP, A) JP-A-SHO 63-131281 (JP, A) Actually opened 63-179564 (JP, U)

Claims (1)

[Claims] 1. A first and a second light source for irradiating light from two different directions to an information carrier on which information is carried by an uneven pattern formed on a surface to be read, and light by the first light source. A reading sensor for optically reading the reflected light from the information carrier when illuminated, and the reflected light from the information carrier when illuminated by the second light source, and the information carrier The reading signal of the reading sensor when light is emitted from the first light source and the reading signal of the reading sensor when the information carrier is illuminated by the second light source are converted into digital signals, respectively. When the digital signal values are Da and Db, and the predetermined threshold value is S, Da and Db are | Da−Db | ≦ S. Condition 2 indicated by Db> S, indicated by Db-Da> S Which of the following three conditions is satisfied, and a ternarization processing means for outputting a ternary signal based on the result of the judgment, and a region which satisfies the above condition 2 based on the ternary signal A first thinning processing means for extracting a first thin line which is a boundary of the concave-convex pattern of the information carrier, and the information carrier of the information carrier from the region satisfying the condition 3 based on the ternary signal. An unevenness characterized by having a second thinning processing means for extracting a second thin line which becomes a boundary of the uneven pattern, and a connection processing means for connecting between the extracted first and second thin lines. Pattern reader.