JPS61253590A - Noise detecting device - Google Patents

Abstract

PURPOSE:To detect noises produced by dust, flaws, etc. of a reading window by putting plural pictures read by a reading means on each other and using the signal produced by putting those pictured on each other. CONSTITUTION:An AND circuit 19 is provided together with a picture buffer 20 and a detecting part 21. The buffer 20 stores the picture information (p0, p1...p9) consisting of the same number of picture element (bits) as the picture read by the 2nd scanning part 13 and having AND secured with the signals of each picture element of the read pictures supplied successively from the circuit 19. Thus 10 sheets of postal matters p0-p9, for example, are read successively by the part 13 and only the black signals having coincident pattern signals of the postal codes are left and put on each other. Thus it is possible to detect noises produced by dust, flaws, etc. of a reading window 22 of the part 13 for said read pictures according to whether the number of black signals of the pictures (pattern signals) put on each other is equal to a prescribed level or not.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention provides a method for automatically reading and sorting postal codes, for example.
The present invention relates to a noise detection device for a read image in a character reading device that reads a postal code written on a transported mail item.

[Technical background of the invention and its problems] Conventionally, automatic postal code reading and sorting machines continuously read postal codes from a plurality of mail items as sorting information, and then read each postal code according to the recognition result of each postal code. It is designed to sort mail.

However, in the above-mentioned apparatus for reading and sorting postal codes, the reading window of the reading section of this apparatus may be coated with glue or dust for sealing the mail, or may be scratched. If there are such stains or scratches, for example, when reading the postal code on a mail item as shown in Figure 7 (a), the read image will look like the one shown in Figure 7 (b), and the postal code will be lost. I accidentally read the number r123J as rl 28J, and again! ! ! ! Unidentifiable (e.g. r12*j
), and the reading performance of the sorting machine may deteriorate.

Conventionally, a read image is displayed on a monitor, and a staff member visually checks the monitor to determine whether there is dust or scratches on the reading window, and the reading window is cleaned or replaced in accordance with this determination. For this reason, the operation by the staff member is troublesome.

[Objective of the Invention] The present invention has been made in view of the above-mentioned points, and an object thereof is to provide a noise detection device capable of detecting the occurrence of noise in a read image due to dirt or scratches in the reading window. be.

[Summary of the Invention] In order to achieve the above object, the present invention reads and scans the information on the object to be read that is sequentially supplied with a reading means, superimposes a plurality of read images read by the reading means, and Noise on the read image is detected based on the signals obtained through superimposition.

[Embodiments of the Invention] Hereinafter, an embodiment of an automatic postal code reading and sorting machine to which a noise detection device according to the present invention is applied will be described with reference to the drawings.

FIG. 2 shows an object to be read according to the present invention, for example, a postal object Pa.
This is an example, and a plurality of lines (two lines in this case) of destination information 1 are written horizontally on the surface. In such a postal item P(l, the postal code (123>3) to be read is written in the top line 2 of the destination information 1. Then, such a postal item Pa is sequentially conveyed in the direction of the arrow a shown in the figure. It has become so.

FIG. 1 schematically shows the configuration of an automatic postal code reading and sorting machine to which a noise detection device according to the present invention is applied. That is, reference numeral 11 denotes a first scanning section, which is provided along a conveyance path along which, for example, the above-mentioned mail items PrJ... are sequentially conveyed. This first scanning unit 11 scans the mail P
Predetermined field of view 4 on a (see Figure 2)
A pattern signal is obtained by optically scanning the image in the direction of the dashed arrow and photoelectrically converting it.
It is composed of a light source that emits light, a self-scanning COD image sensor that receives the reflected light, and converts it into an electrical signal.

Reference numeral 12 denotes a character line detection device, which detects the position of the top line 2 where the postal code 3 to be read exists based on the pattern signal from the first scanning unit 11, and sets the reading field position of the second scanning unit 13, which will be described later. For example, an X mask creation circuit that creates an X mask for projection in the X direction and a Y mask for creation of a Y mask for projection in the Y direction based on the supplied pattern signal. The creation circuit and each line information and each condition information extracted by these mask signals determine whether or not zip code 3 in the X mask is suitable as a zip code block, and the X on the left side of the determined block is It is composed of a determination circuit, etc. that selectively outputs the coordinate value and the Y coordinate value of the upper left end.

Reference numeral 13 denotes a second scanning section, which is provided in front of the first scanning section 11 in the conveyance direction along the conveyance path of the mail P(+). As shown in the figure, a pattern signal is obtained by optically scanning the set visual field range 5 on the postal item PO, that is, the position where the postal code (123>3) is written and photoelectrically converting it.
Ru. The second scanning unit 13 includes a light source (not shown) that irradiates light onto the mail Pa through a reading window 22 made of glass, and a reading field of view that is controlled by the output from the character line detection device 12. A self-scanning COD image sensor 23 that receives the reflected light from the postal item Pa via the reading window 22 and the movable mirror and converts it into an electrical signal. be done.

Reference numeral 14 denotes a character detection/cutout section which detects and cuts out the postal code 3 in the top row 2 of the destination information 1 based on the pattern signal from the second scanning section 13. For example, a mask is created by the character detection and cutting section 14, each mask is divided into blocks, and each character pattern of a postal code block is detected and cut out one character at a time from this block pattern.

A normalization section 15 normalizes and samples each character pattern supplied from the character detection and extraction section 14, that is, each detected and extracted number pattern of the postal code.

Reference numeral 16 denotes a recognition unit which converts the postal code 3 supplied from the character detection and extraction unit 14 through the normalization unit 15 into a dictionary 17, for example.
By matching the numbers in the box with the reference pattern! ! ! The recognition result is output to a subsequent sorting machine (not shown).

On the other hand, a noise detection device 18 is connected to the second scanning section 13 . This noise detection device 18 reads a plurality of mail pieces, for example, 10 pieces of mail pieces Pa to P9 sequentially by the second scanning unit 13, and leaves only the black signal that matches the pattern signal of each postal code. They are sequentially superimposed, and depending on whether the number of black signals in the superimposed images (pattern signals) reaches a specified value, noise caused by dirt or scratches in the reading window 22 of the second scanning unit 13 is removed from the read image. It is something to detect. As shown in FIGS. 5(a) to 5(d), the noise detection device 18 is sequentially supplied with read images (postal code pattern signals) po-Di that are continuously read from the second scanning unit 13. AND circuit 19, image information consisting of the same number of pixels (bits) as the image read by the second scanning section 13, and obtained by ANDing the signals of each pixel in the read image sequentially supplied from the AND circuit 19. (1) on ptn p2n
Image buffer memory 20 for storing °・°np9)
1 and the number of pixels S of the noise information N appearing in the image finally stored in the image buffer memory 20, for example, the image p shown in FIG.
an pln D2n...npg),
It is constituted by a detection section 21 that detects noise depending on whether the number S of black pixels has reached a specified value. The above specified value is a predetermined threshold value (TH>),
The number of pixels in the vertical direction (H) and the number of pixels in the horizontal direction (W
> (S≧T old XHXW).

The output of the noise detection device 18, that is, the noise detection signal output from the detection section 21, is supplied to a subsequent sorting machine (not shown). As a result, the sorting machine detects the occurrence of noise in response to the noise detection signal by lighting up a noise alarm indicator (not shown) provided on the operation panel.
This will notify the staff. Furthermore, the 10 pieces of mail used for noise detection at this time are sorted into a reject collection section, and the sorting operation is stopped.

Next, the operation in such a configuration will be explained.

For example, now, the mail PO shown in FIG. 1 is conveyed in the direction of arrow a and reaches the first scanning section 11. Then, the first scanning unit 11 optically scans the field of view 4 on the mail Pa and performs photoelectric conversion, and the pattern signal is supplied to the character line detection device 12 to detect this character line. Device 1
2 detects the position of the top row 2 where the postal code 3 to be read exists, converts it into a reading field position setting signal for the second scanning section 13, and outputs it. Thereby, in the second scanning section 13, the movable mirror for the reading field position is controlled by the field position setting signal, so that the field of view range 5 on the mail Pa is set. When the mail Pa reaches the second scanning section 13, the second scanning section 13, for example, as shown in FIGS. 3 and 4,
The visual field range 5 on the mail Pa is optically scanned and photoelectrically converted, and the pattern signal is output to the character detection cutout section 14 and the noise detection device 18. This character detection cutting section 1
4 detects and cuts out, for example, the block of postal code 3 in the top row 2 of the destination information 1 based on the pattern signal from the second scanning unit 13. That is, the detection/cutout section 14 creates a mask in the block of the postal code 3, blocks each mask, and detects and cuts out character patterns one character at a time from each block pattern.

Then, the character pattern of this postal code 3 is
Sent to 5. This normalization unit 15 normalizes each character pattern of the above-mentioned postal code 3, samples it, and sends it to the recognition unit 1.
Output to 6. Then, the recognition unit 16 converts the character patterns supplied from the character detection and extraction unit 14 via the normalization unit 15, that is, each numeric pattern of the postal code 3, into a standard corresponding to the digits in the digit number 17, for example. Each character is recognized by a method such as matching with a pattern, and the postal code as a result of the recognition is output to a subsequent sorting machine (not shown). Further, the pattern signal from the second scanning section 13 is supplied to the noise detection device 18 as an image po as shown in FIG. 5(a), for example. The image (postal code pattern signal) po is sent to the uniformity buffer memory 20 via the AND circuit 19, and the read image po is stored as is in the image buffer memory 2G.

Then, the mail P1 (not shown) is conveyed in the direction of arrow a following the mail PO, and is transported to the first scanning unit 1.
Reach 1. Then, the first scanning unit 11 optically scans the field of view on the mail Pl in the same manner as described above, detects the position where the postal code to be read by the character line detection device 12 exists, and reads the postal code. A visual field position setting signal is output to the second scanning section 13. As a result, the second scanning section 13
optically scans the field of view on the mail Pl, and outputs the pattern signal to the character detection cutout section 14 and the noise detection device 18. This character detection cutting section 14
Based on the pattern signal from the scanning section 13, each numerical pattern is detected and cut out one character at a time, and outputted to the recognition section 1B via the normalization section 15. Then, the recognition unit 16 performs character recognition for each number pattern of the postal code by, for example, a matching method with a reference pattern corresponding to the number in the dictionary 17, and sends the result to a subsequent sorting machine (not shown).
Output to. Also, the mail P from the second scanning section 13 is
1 pattern signal, each pixel of which is sequentially detected by the noise detection device 18 as an image p1 as shown in FIG. 5(b), for example.
The signal is supplied to the AND circuit 19 within. Further, at this time, the content in the image buffer memory 20, that is, the signal of each pixel of the image po of the postal item Pa is also sequentially supplied to the AND circuit 19. As a result, the AND circuit 19 outputs the image p1 of the postal item P1 supplied from the second scanning unit 13 and the image p1 of the postal item Pa stored in the image buffer memory 20.
Only the image (pon E) t) obtained by logical AND with o is supplied to the image buffer memory 2G. As a result, in the image buffer memory 20, there remains an image of only the matching portion of images pa and pl, that is, an image in which only matching black signals remain in each pixel! 1 (pon pt
) is memorized.

Next, the postal code of mail P2, which is conveyed on the conveyance path following mail P1, is read and recognized, and the result is output to the subsequent sorting machine, and the read image p2 is output to the noise detection device 18. . Then, this noise detection device 1
In step 8, the image (E)on pt) of the postal items Pa and Pl subjected to the logical product is read out from the image buffer memory 20, and this image (Den pl) and the postal item P2 read by the second scanning unit 13 are read out. The image p2 (not shown) is supplied to the AND circuit 19. Then, from this AND circuit 19, an image (pan pl) obtained by taking the AND of the postal item Pa and Pl and an image p of the postal item P2 are output.
Image obtained by ANDing with 2 (pen pln p2
) are supplied to the image buffer memory 20. As a result, the image buffer memory 2G contains images 1) o,
Only the black signal of the matching portion of pl and p2 is stored.

In the same manner as above, the postal codes of each of the mail items P3 to P9 that are sequentially conveyed on the conveyance path following the mail item P2 are read and recognized, and the results are output to the subsequent sorting machine, and the read images p3 to P9 are respectively output. pg is sequential noise detection device 1
8, each image p3 to p9 is superimposed to create an image of only the matching part (pen ptn ptn...
-rle> is stored in the image buffer memory 20.

Then, in the image buffer memory 20, a logical product image (Den D
ln D2n °°°nps) is stored,
The image information is supplied to the detection section 21. Then, the detection unit 21 determines the number of black pixels S based on the number of black signals in the supplied image information, and detects noise based on whether the number S of black pixels reaches a specified value. When the number S of black pixels reaches a specified value, it is assumed that noise has been detected, and a noise detection signal is output to a sorter (not shown). As a result, the sorting machine notifies the staff of the occurrence of noise by lighting up a display (not shown), and also
-Ps is accumulated in the reject accumulation section and the sorting operation is stopped. Through this detection, the staff member knows that a dent or scratch has occurred in the reading window 22 of the second scanning section 13, and cleans or replaces the reading window 22. As a result, the subsequent sorting machine can prevent erroneous sorting due to the above-mentioned noise.

Further, when the noise detection signal is not output from the detection unit 21, the sorting machine continues the sorting operation.

As mentioned above, the number of black signals common to all images, that is, the number of black pixels, is calculated by taking the logical product of the reading surface images for 10 pieces of mail, and this number of black pixels is specified as Noise is detected based on whether the value has been reached. This makes it possible to detect noise caused by dirt, scratches, etc. in the reading window of the reading unit.

In the above embodiment, 10 pieces of mail were used as an example to explain the operation of the noise detection device, but the present invention is not limited to this, and similar detection may be performed using multiple pieces of mail other than 10 pieces. be able to.

However, the detection results must be output before the first read mail is sorted by the sorter. In addition, although it has been explained that noise is detected from the number of black signals of the image information obtained by taking the logical product of multiple read images, this is not limited to this, and the image is obtained by inverting the multiple read images. The noise may be detected from the number of black signals of the image information obtained by calculating the logical sum of the logical sum and inverting the image obtained by calculating the logical sum.

[Effects of the Invention] As described above, according to the present invention, it is possible to provide a noise detection device that can detect noise generated in a read image due to dust, scratches, etc. on the reading window of the reading unit.

[Brief explanation of the drawing]

FIG. 1 to FIG. 6 show an embodiment of this invention.
Fig. 1 is a block diagram schematically showing the configuration of an automatic postal code reading and sorting machine, Fig. 2 is a plan view showing an example of mail, and Fig. 3 is a block diagram schematically showing the configuration of an automatic postal code reading and sorting machine.
4 and 4 are diagrams showing the configuration of the reading section, FIGS. 5 and 6 are diagrams showing examples of read images, and FIGS. 7(a) and (b) are reading targets and reading images in the conventional example. FIG. 3 is a diagram for explaining an image. Pa-... Mail, 3... Postal code, 5... Setting visual field range, 11... First scanning unit, 12... Character line detection device, 13... Second scanning unit, 14...Character detection cutting unit, 16...Recognition unit, 17...Dictionary, 18...
Noise detection device, 19...AND circuit, 20...Image buffer 1 memory, 21...Detection section, 22...Reading window, 23...COD image sensor. Applicant's agent Patent attorney Takehiko Suzue s2 figure 1! 3 figure s4 figure s6 figure s7 figure

Claims (1)

[Claims] In an apparatus for reading information on an object to be read that is sequentially supplied as image information, a reading means for reading and scanning the information on the object to be read, a means for superimposing a plurality of read images read by the reading means, and this means. 1. A noise detection device comprising a detection means for detecting noise on a read image based on a signal obtained by the method.