JPH0973512A - Mark reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the recognition rate of a mark in which output is deteriorated owing to a base without detecting noise and the like by using plural thresholds in accordance with the state of the base of a mark printing part and the output level of a sensor. SOLUTION: In a bar code reader, for example, a picture read part 12 emits infrared light on a mail 1 traveling a transportation line from a light source 2, reflected light is received by the sensor 4, the output is A/D converted and is synchronized by the latch circuit 7 of a picture recognition part 13 and it is stored in a picture memory 8 as a multivalue picture. The picture is shifted to a main memory 10 by CPU 9 and CPU 9 processes the picture and encodes it in accordance with a recognition program stored in the main memory 10. Here, CPU 9 uses plural different values in accordance with the state of the base where the mark is printed and the output level of the sensor 4 as the threshold detecting the presence or absence of the mark at the time of processing the picture.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mark reading device for reading a printed mark with an optical sensor, and more particularly, the mark is printed on a medium such as a postal matter whose base color cannot be specified and is used for reading output. The present invention relates to a mark reading device which has a large variation and is difficult to recognize.

[0002]

2. Description of the Related Art In a mark reading apparatus, an optical signal from a mark is converted into an electric signal by a photoelectric conversion element or the like, and the voltage is compared with a threshold value to determine whether the voltage is higher than the threshold value or not. Has been decided. In the conventional mark detection method, for example, as described in Japanese Patent Application Laid-Open No. 55-99677, in the method of detecting a mark by using a plurality of threshold values, two threshold values are used. There is a method in which a mark is detected for each value, and if the obtained result is the same, that is adopted as the result, and if the obtained result is different, the mark cannot be determined. In the above detection method, the entire detection medium is detected with a fixed threshold, so in the case of marks printed on media such as mail that cannot be identified as the background, the output may vary depending on the background color (address writing, etc.). Therefore, there is a problem that an undetectable part occurs.

[0003]

Ink (hereinafter abbreviated as ink) that has been conventionally excited by infrared light and emits light in the infrared wavelength region has a variation in output depending on the background color. Therefore, when the entire multi-valued image obtained by the image reading apparatus is determined from the output change amount by setting a fixed threshold value, there is a portion where the mark shape cannot be recognized. In particular, the output is remarkably reduced for characters written with a writing instrument having a high carbon content or printed matter. As a result, there is a high possibility that the output will be reduced for mail items whose base color cannot be specified. FIG. 2 is a diagram showing a printed diagram of a 4-state barcode with a background character, a main scanning line output diagram, and a coordinate image diagram of threshold values S1 and S2. Here, the white rectangle is the barcode, and the black character is the character written on the background. The image output in the main scanning direction at the position 101 in FIG. 2A is as shown in FIG. 2B. Figure 2 (b)
As is clear from the above, the plain area 102 and the area 103 where the characters are overlapped have extremely different outputs. After all, when the bar code of FIG. 2A is output using the threshold value S1, the result becomes as shown in FIG. 2C, and when the bar code of FIG. As shown in d). As shown at 104 in FIG. 2C, the code output of the portion overlapping the character is detected in a shape different from the actually printed mark. In order to recognize the position of the bar where the characters are overlapped, it is necessary to set the threshold value to the threshold value S2 of FIG. However, in that case, the ink bleed portion 105 between the marks shown in FIG.
Since the ink splatter 106 and the like at the time of printing have almost the same output as that of the output lowering portion, there is an adverse effect of being erroneously detected as a mark, and it is difficult to detect the output lowering portion. An object of the present invention is to solve such a conventional problem and to detect an output reduction part due to a background without detecting noise or ink bleeding part, and to improve the recognition rate. It is to provide a mark reading device.

[0004]

In order to achieve the above object, in the mark reading apparatus of the present invention, the state of the base on which the mark is printed at the threshold value for detecting the presence or absence of the mark,
Also, by using a plurality of different values depending on the output level of the sensor, a mark with a low output is detected without being affected by a print bleed portion, noise, and the like. The ink output has a high threshold value (S) as shown in FIG.
If the amount of change is detected in 1), printing bleeding and ink splattering are not visible as shown in FIG. In the present invention, paying attention to this, an area (width, height) 109 having a mark is specified in advance with a high threshold value, and only the area within that area has a low threshold value. By using S2 for detection, it is possible to detect the position of the mark even in the area where the output is reduced due to the background.

[0005]

In the present invention, the difference in the ink output characteristics due to the base is absorbed by using two threshold values in one detection. One is a threshold value for noise removal set to a relatively high level, and the other is a threshold value for detection of a reduced output part set to a low level. In the present invention, by combining these, it is possible to detect the output lowering portion. As a result, it is possible to detect the output lowering portion due to the background without detecting noise or ink bleeding portion, and improve the recognition rate.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings by taking a bar code detection which is a kind of mark as an example.
FIG. 1 is a configuration diagram of a bar code reader according to an embodiment of the present invention. The mark reading device of the present invention is composed of two blocks, which are an image reading unit 12 and an image recognition unit 13. The image reading unit 12 irradiates the postal matter 1 running on the conveyance path with infrared light from the light source 2, magnifies the reflected light from the printed bar code with the lens 3, and receives it with the sensor 4 including a plurality of photodiodes. . The output of the photodiode is amplified by the amplifier 5 and converted into a digital signal by the AD converter 6. The converted signal is passed to the image recognition unit 13, synchronized with the latch circuit 7, and then stored in the image memory 8 as a multi-valued image. The stored image is transferred to the main memory 10 by the CPU 9 via the general-purpose bus, and the CPU 9 processes and encodes the image according to the recognition program also stored in the main memory 10. The coded data is passed to the upper layer via the serial interface (serial driver) 11. According to the present invention, when the CPU 9 processes an image, a plurality of different values are used according to the state of the background on which the mark is printed on the threshold value for detecting the presence or absence of the mark and the output level of the sensor. The low output mark can be detected without being affected by the print bleeding portion, noise, and the like.

The preprocessing of bar code encoding according to the present invention will be described below with reference to the case where the medium is a postal matter. FIG. 2A is a 4-state bar code printed diagram with base characters, FIG. 2B is an output diagram of the main scanning line in FIG. 2A, and FIG. 2C is FIG. 2B. 2D is a coordinate image diagram of the threshold value S1 in FIG. 2, and FIG. 2D is a coordinate image diagram of the threshold value S2. As described above, in the coordinate image detected with the threshold value S1, 104 in FIG.
As shown in, the bar code that overlaps the character cannot be detected. On the other hand, in the coordinate image detected at the threshold value S2,
As shown by 105 and 106 in FIG. 2D, the bleeding portion between the marks and the ink splattering portion are detected, and the accurate barcode cannot be detected. Therefore, in the present embodiment, in order to detect a mark that has a low read output and is difficult to recognize, two different thresholds S1 and S2 are used to detect a low level output without being affected by noise. To do.

FIG. 3A shows a coordinate image rectangle detection output, and FIG. 3B shows a coordinate image bar width detection output.
3C is a coordinate image bar position specifying output, which is shown in FIG.
Is a bar specific position cutout image output. In the present embodiment, the outline 107 of the entire mark shown in FIG. 3A and the width 108 of each mark shown in FIG. 3B are specified from the coordinate table 1 detected by the threshold value S1 that is not affected by noise. Then, the mark print area 109 is created. This area 1
By checking the inside of 09 with a threshold value S2 that can be detected even with a low output, the image of FIG. 3C is output, and the coordinate table 2 is created from this. A mark as shown in FIG. 3D is output from the coordinate table 2 and coded.

FIG. 4 is a flow chart of a mark reading operation showing an embodiment of the present invention, and FIG. 5 is a format diagram of the coordinate tables 1 and 2 used in the present invention.
The processing procedure of the present invention will be described with reference to FIG. 4 while referring to FIGS. 2, 3, and 5. 1) First, a multi-valued image is checked using a threshold value S1 in the direction of the arrow 101 (main scanning direction) shown in FIG. 2A, and the output is above the threshold value S1 shown in FIG. 2B. The first changing point that rises is the starting point of the bar, and the last changing point that falls is the ending point (steps 41 and 4 in FIG. 4).
2). In FIG. 2B, 102 is a plain output, 10
3 is an output of the character overlapping portion. 2) Next, the above check 1) is sequentially repeated from the left end in the sub-scanning direction to create the coordinate table 1 of the start point and the end point (step 43 in FIG. 4). An image connecting the start point and the end point is shown in FIG. When detected with the threshold value S1, the shape becomes different from the actual mark, as indicated by 104 in FIG. 2C. If the last line is detected, the next step is performed (step 44 in FIG. 4). The starting point and ending point coordinate table 1 has a format and contents as shown in FIG. Coordinate table 2 described later
Is the same as the coordinate table 1. In the coordinate table, the starting point coordinates, the ending point coordinates (the number of pixels in the main scanning direction), and the line positions (the number of pixels in the sub scanning direction) of the lines that form the bar are arranged in the sub scanning direction in this order. Here, an explanation from the first bar start point 1 and the end point 1 to the second bar start point 5 and the end point 5 is shown in FIG. Here, 2
Only the main item is stored, but in the case of FIG.
Up to the start point 5 and end point 5 of the seventh bar will be stored.

3) Next, the rectangular area of the entire bar code as shown by 107 in FIG. 3A is specified from the coordinate table 1 created in 2) above from the maximum bar length and the outermost bar coordinates ( Step 45). Here, the outermost bar coordinates are
It is the position of the start point end point line that occurs first and last when viewed from the sub-scanning direction shown in FIG. 2A, that is, the position (coordinates) of the outer lines of the bars at both ends of FIG. 3A. 4) Next, from the coordinate table 1 created in 2) above, 108 of the effective width of each bar in FIG. 3B is specified (FIG. 4).
Step 46). At this time, if the bar width is smaller than the original bar width, the bar width at the time of printing is set as the area width. 5) Next, the detection area indicated by 109 in FIG. 3C is specified from the rectangular area created in 3) above and the bar width specified in 4) above (step 47 in FIG. 4). 6) Next, in the same method as 1) above, the area range specified in 5) above is checked with the threshold value S2 shown in FIG. 2B (step 48 in FIG. 4), and in FIG. The coordinate table 2 is created so that it becomes an image of
9). When the final area is detected (step 50 in FIG. 4), the coordinate table 2 is used for coding (step 51 in FIG. 4).

By using the coordinate table 2 created in this embodiment for the character overlap, there is no effect on ink bleeding, noise, etc., and it is possible to detect the bar, which facilitates coding. Become. Further, in the present embodiment, the method of creating the bar coordinates from the output change points of the multi-valued image and detecting the position has been described. However, in addition to this, the output is sliced by the threshold values of S1 and S2, and the binary value It is also effective to create a coded image, specify an area from the binarized image of S1 as in the present embodiment, and cut out only the specific area from the binarized image of S2 to form a coded image. In the present embodiment, in particular, the threshold value for detecting a mark printed with an ink containing a phosphor that is excited by infrared light and emits light in the infrared wavelength region, depending on the state of the base on which the mark is printed. By using a plurality of different values, it is possible to read the presence or absence of a mark having a low output without being affected by noise.

[0012]

As described above, according to the present invention,
Since multiple thresholds are used to detect the position for mark recognition, it is possible to detect the area where the output is reduced due to the background without being affected by noise or ink bleeding. It is possible to improve.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a bar code reader to which the present invention is applied.

FIGS. 2A and 2B are a 4-state bar code print diagram including a background character, a main scanning line output diagram, a coordinate image diagram of a threshold value S1, and a coordinate image diagram of a threshold value S2.

FIG. 3 is a diagram of a coordinate image rectangle detection diagram, a coordinate image bar width detection diagram, a coordinate image bar position specification diagram, and a bar specific position cutout image in the present invention.

FIG. 4 is an operation flowchart of a mark reading device showing an embodiment of the present invention.

FIG. 5 is a format diagram of coordinate tables 1 and 2 used in the present invention.

[Explanation of symbols]

1 ... Mail, 2 ... Light source, 3 ... Lens, 4 ... Sensor,
5 ... Amp, 6 ... AD converter, 7 ... Latch circuit, 8 ...
..Image memory, 9 ... CPU, 10 ... Main memory, 11 ...
-Serial driver, 12 ... Image reading unit, 13 ... Image recognition unit, 101 ... Main scanning direction output detection line, 10
2 ... Plain output, 103 ... Character overlapping output, 104 ...
Character overlapping image, 105 ... Ink bleeding detection spot, 106 ... Print skipping spot, 107 ... Bar code rectangle, 108 ... Bar width, 109 ... Bar detection area.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Yoshinari Aoshima 1-88, Toyora, Ibaraki-shi, Osaka Inside Hitachi Maxell Co., Ltd.

Claims (3)

[Claims] 1. A mark reading device for reading a printed mark with an optical sensor, wherein a plurality of different values are used as a threshold value for detecting the presence or absence of the mark, depending on the state of the substrate on which the mark is printed and the output level of the sensor. A mark reading device comprising an image recognition unit capable of detecting a mark having a low output without being affected by a bleeding portion of printing or noise by using a value. 2. The mark reading device according to claim 1, wherein the image recognition unit specifies the position of the mark in the image read with a high threshold value that is not affected by noise and the like, and then specifies the mark position. A mark reading device characterized by reading the presence / absence of a mark with a threshold value having a low level with respect to a predetermined position. 3. A mark reading device printed with an ink containing a phosphor that is excited by infrared light and emits light in the infrared wavelength region. A substrate on which the mark is printed as a threshold value for detecting the mark. By using a plurality of different values according to the state of the sensor and the output level of the sensor, an image recognition unit that can detect a low output mark without being affected by a print bleed portion or noise is characterized. Mark reading device.