JPH08329185A - Symbol information reader - Google Patents

Abstract

PURPOSE: To provide a symbol information reader capable of quickly and surely reading out bar code symbol information from a moving two-dimensional bar code. CONSTITUTION: The symbol information reader is constituted of a stuck bar code label 2 with a PDF-417 format e.g. stuck to an article 1 or the like moved by a carrying system 3 such as a belt conveyer, an image pickup optical system 4 for optically forming an image of the label 2, a two-dimensional image pickup part 5 for reading out the label 2 formed as the image by a photoelectric conversion face, a frame memory 6 for temporarily storing the read label picture data (video signal), and a data processor 7 for decoding the label picture data into original symbol information and allowed to read out the symbol information of a moving bar code symbol.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a symbol information reader for reading symbol information such as bar codes.

[0002]

2. Description of the Related Art Conventionally, printed on articles, or
There is a symbol called a barcode symbol, which is printed on the attached tag or the like and indicates information about the article. This bar code symbol is used in a point-of-sale (POS) system which is becoming popular, and has become widely known.

With this bar code symbol, bars having different widths and spaces are combined in parallel to form a bar code character consisting of one pattern. If necessary, there is also a symbol in which a necessary character group including a check digit is arranged in parallel and a characteristic predetermined pattern such as a start / stop character is arranged at the front and rear.

The bar code used for general consumer goods is JAN (Japanese Article Number).
r) is standardized. Another application of barcodes is in logistics symbols. This physical distribution symbol has a 1-digit or 2-digit physical distribution identification code added before the JAN code.

Any of the above-mentioned bar code symbols is called a one-dimensional bar code, and the amount of information that these code systems can have is about several tens of bytes, which is not suitable for actual use. It was enough.

In response to the demand for the information amount of such a bar code, a symbol system called a two-dimensional bar code having a different structure has been proposed. Each of these symbol systems has a characteristic that much more information can be encoded than a one-dimensional barcode. These systems are called stacked barcodes and are a method of increasing the amount of information by stacking one-dimensional barcodes in the direction of the bar. A representative of this stacked bar code is a code system called PDF-417.

Conventionally, as a symbol information reader for reading the stacked bar code, for example, Japanese Unexamined Patent Publication No. Hei.
There is known a laser scanning type reading device as disclosed in Japanese Patent Publication No.-268382. In this reading device, a predetermined bar code is read and symbol information is decoded by projecting and scanning laser light in two dimensions.

Further, Japanese Laid-Open Patent Publication No. 2-268383 discloses an apparatus for picking up a barcode with a two-dimensional image pickup apparatus, loading the image of the barcode into a memory, and decoding the barcode symbol information based on this data. It is disclosed.

Further, such a bar code is converted into FA (F
There is an increasing demand for use as a data carrier in image automation.
This is for recognizing the product name, destination, etc. of the distributed product while the distributed product is moving on a belt conveyor or the like.

Conventionally, R is generally used as a data carrier.
Although an F tag or the like has been used, it is relatively expensive as a tag, and if this can be replaced with a bar code, a low cost data carrier can be realized.

As a conventional information reading device for moving articles, Japanese Patent Laid-Open No. 56-118177 discloses an optical inspection for performing a predetermined inspection based on a video signal obtained by scanning an image of a moving object. An apparatus is described and a circuit is disclosed for detecting from an video signal that an object has passed within an examination field of view.

[0012]

However, the problem with the above-mentioned Japanese Patent Laid-Open No. 56-118177 is that the existence of the object is recognized from the width data of the video signal.
Although it is possible to recognize a one-dimensional barcode such as JAN, in the two-dimensional stacked barcode such as PDF417, an incomplete barcode cannot be obtained even if the barcode label is not completely within the inspection field of view. There is a risk of recognizing the label and going to the reading process. In such a case, the bar code information cannot be read, resulting in a decrease in reading reliability and reading speed.

Further, in order to read the information of the two-dimensional stacked bar code attached to the moving article, a frame image is picked up by using an interlaced image pickup device, and the position read next is read from the position read first. Since it has been moved up to, there is a problem that when two field images are combined, a displacement of field components occurs. Therefore, it is an object of the present invention to provide a symbol information reading device that reliably and quickly reads barcode symbol information of a moving two-dimensional barcode.

[0014]

In order to achieve the above object, the present invention provides an image pickup means for picking up a two-dimensional image of a bar code which is composed of a bar and a space and which represents arbitrary information as a pattern. Storage means for temporarily storing sequentially obtained two-dimensional images, transfer means for transferring the two-dimensional images from the image pickup means to the storage means, and two-dimensional images stored in the storage means are sequentially read and imaged. Recognition means for recognizing the presence / absence of a barcode symbol within the range, and position detection for detecting the presence position and inclination of the barcode recognized by the recognition means within the imaging range to detect the scanning direction of the barcode. A reading means for sequentially reading information of the barcode from a two-dimensional image of the barcode according to the scanning direction detected by the position detecting means; From information keycode, it provides symbol information reading apparatus constituted by a decoding means for decoding the arbitrary information.

The transfer means provides a symbol information reader for compressing and transferring the two-dimensional image picked up by the image pickup means. Further, the transfer means transfers one scan line of the field component of the two-dimensional image picked up by the image pickup means, vacates one line of the storage means, and transfers the scan line again to obtain one field of image information. A symbol information reader for transferring is provided, which includes a processing operation of repeating the process until transferring, and copying the image information of a line one line above or below that line to a vacant line after the transfer.

[0016]

The symbol information reading apparatus having the above-described structure captures a moving two-dimensional bar code as a two-dimensional image and temporarily performs a predetermined compression process on the obtained two-dimensional image. Memory, transfer the captured image to the storage means,
After automatically recognizing the presence or absence of the barcode symbol, the position of the barcode symbol is recognized, the barcode information is sequentially read, and the barcode information is decoded into the original information.

[0017]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 shows a schematic configuration of a symbol information reading apparatus as a first embodiment, which will be described. This symbol information reading device is a configuration example for reading the symbol information of a bar code symbol attached to a moving article or the like. Here, in the present embodiment, the symbol information is, for example, information about an article, and is, for example, an article name, a destination, or the like. Of course, the symbol information depends on what is used.

This symbol information reading device is a stacked bar code label (hereinafter referred to as a bar code label) of, for example, PDF-417 format attached to an article 1 or the like which is moved by a conveyor system 3 such as a belt conveyor.
2, an image pickup optical system 4 that optically forms an image of the barcode label 2, a two-dimensional image pickup unit 5 that reads the imaged barcode label 2 on a photoelectric conversion surface, and read label image data (video signal). Frame memory 6 for temporarily storing
And a data processing device 7 for decoding the label image data into the original symbol information. Further, a host device 16 such as a microcomputer for processing symbol information is provided outside. Further, label image data (frame image) read by the two-dimensional image pickup unit 5
In addition, an image transfer unit 17 that performs image processing such as image compression and stores the processed image data in the frame memory 6 may be provided.

The data processing device 7 includes a label detecting section 8 for detecting whether or not the label image data stored in the frame memory 6 is readable within the photographing range of the image pickup optical system 4. Barcode label 2
A reading unit 9 that scans (label image data) and reads it as label information, a decoding determination unit 10 that determines whether the read label information can be decoded, and the label information that is determined to be decoded is used as the original symbol information. It is composed of a decoding unit 11 for restoration, a memory (including various registers for storing various constants and variables) 12, and a control unit 13 for controlling each of these parts.

The image pickup optical system 4 and the two-dimensional image pickup section 5 are, for example, CCD cameras. Further, the label detection unit 8
Is a position detecting unit 14 for detecting whether the label image data stored in the frame memory 6 is within the photographing range of the photographing optical system 4, that is, the position where the barcode label 2 photographed within the photographing range is not chipped. And an inclination detecting section 15 for detecting the inclination of the barcode label 2.

The data processing unit 7 can be replaced with, for example, a personal computer including a CPU and a memory. Next, the stacked bar code described above will be described.

FIG. 2 shows a label structure of PDF-417 as an example of a stacked bar code. The bar code label 2 is a label portion 21 which is an area of an information component to be decoded which is composed of a bar code character group consisting of a combination of a bar and a space, and a start / stop character which is arranged before and after the label part 21. Code 2
2 and a stop code 23. And 1
The code consists of four bars and a space except for the stop code 23. The start and stop codes 22 and 23 start from large bars 22A and 23A called "big bars".

The label portion 21 has a start code 22.
And a plurality of data columns 21 in which a code called a row indicator 21A existing next to the stop code 23 and the actual data sandwiched between them are described.
And a label matrix 21C of B. The row indicator 21A describes the label size in the row and column directions, the security level, and the like. Therefore, by decoding the information of this row indicator, the information size of the label, the number of recoverable codes, etc. can be determined.

FIG. 2 shows a bar code label having a 4 × 2 label matrix. Label detection, label information reading, and decoding in the data processing device 7 will be described with reference to the flowchart shown in FIG. FIG. 3 shows a schematic diagram in which the label image of PDF-417 as described above is virtually projected on the pixel array of the frame memory 6.

First, various parameters such as a frame memory, a threshold value and various global variables are initialized (step S1). Next, an image capturing routine, which will be described later, is called to capture the label image data (video signal) into the frame memory 6 (step S2). Next, a label detection routine, which will be described later, is called, it is checked whether a bar code label is present using the captured label image data, and if the bar code label is present, the label image data is detected ( Step S3).

Then, the result of the label detecting process in the step S3 is judged (step S4), and if there is no label (NO), the process proceeds to the step S1 again, and the image capturing process is executed again. I do.

Next, if the label exists (YE
S), by calling a threshold value determination routine to be described later, in the embedding check in step S11, the optimum scan routine in step S12, and the intelligent scan routine in step S15, the width between the line data to be processed and the edge is processed. A threshold value (variable THRESHOLD) used in the process for extracting information is obtained (step S5).

Next, the indicator information determining routine is called to read the row indicator 21A of the label 2 to determine the label size and the like (step S6). Then, it is determined whether the label size or the like is determined in the determination routine of step S6 (step S6).
7) If not determined (NO), the process proceeds to step S1 again, and the image capturing process is performed again. On the other hand, when the label size and the like are determined (YES),
A tilt redefinition routine, which will be described later, is called to obtain a more accurate label tilt (step S8).

Next, a scan equation determination routine, which will be described later, is called to define various variables for scanning the entire label 2 (step S9). Next, an embedding check routine, which will be described later, is called to estimate the position of each codeword, and it is further checked whether the position of each codeword is within the screen (step S10).

Then, in step S10, it is judged whether or not the number of codes estimated to be unreadable exceeds the number that can be repaired (step S11). If the number exceeds the recoverable number, decoding is impossible (NO). Then, the process proceeds to step S1 again, and the image capturing process is performed again. On the other hand, if it is below the threshold, decoding is possible (YES), an optimum scan routine as described later is called, and the above-mentioned step S9 is executed.
The label is scanned over the entire surface at the optimum intervals by using the various variables defined in step S12, and the label information is read (step S12).
Here, the optimum scan refers to a scan at an optimum interval that requires the least amount of calculation and allows all label information to be determined.

Then, it is judged whether or not the information read by the optimum scan in step S12 is decodable (step S13). If it is decodable (YES), the process proceeds to step S16 to perform the decoding process. But,
If decoding is not possible (NO), an intelligent scan routine, which will be described later, is called, and step S10 is executed.
Using the various variables defined in step S12,
The code position that could not be read in is scanned and the label information is read (step S14).

Next, it is judged whether or not the information read in step S14 is decodable (step S15), and if it is decodable (YES) and if the information read in step S12 is decodable, a decoding process is performed. Is performed (step S16). After the completion of this decoding process, the process returns to step S1 to read a new barcode label.

However, if the judgment in step S16 indicates that decoding is not possible (NO), the process moves to step S1 again to capture an image. In the day code process of step S16, the read information is decoded by the optimum scan routine and the intelligent scan routine, and the decoded symbol information is output to the host device 14.

Next, various processing routines in the flowchart of FIG. 4 will be described in detail below. First, the image capturing routine called in step S2 shown in FIG. 4 will be described with reference to the diagram shown in FIG. 5 for capturing only the field screen and the flowchart shown in FIG.

The image capturing routine includes a process of transferring the label image data (video signal) captured by the two-dimensional image capturing section 5 and temporarily storing it in the frame memory 6.

When the interlaced two-dimensional image pickup section 5 is used to read the bar code label attached to the moving article, it is usually impossible to read the bar code information from the frame screen. That is, as described in the problems of the related art, in the interlaced two-dimensional imaging device 5, two field screens (an odd component and an even component) for scanning an image while skipping one line at a time are combined to form one frame screen. However, since the barcode label to be imaged is constantly moving,
This is because there is a positional shift between the scanning of the second field image and the scanning of the second field image, and when the two field images are combined, the barcode label image is misaligned.

Therefore, in this embodiment, when the moving bar code label is read by using the interlace type two-dimensional image pickup device 5, the image information of only the field screen is transferred and taken in.

In the image capturing routine, either (even component or odd component) field screen is transferred and captured (step S21). In this case, the field image is fetched in the frame memory 6 without leaving a row, and the contents of the frame memory 6 at that time are as shown in FIG. In FIG. 5, the field image may have either an odd component or an even component. After this processing ends, the process returns to the main flowchart (FIG. 4).

However, in this case, one field image is compressed vertically as shown in FIG.
The height per line of the bar code label is reduced,
The reading conditions are stricter than the frame image. Therefore,
This method can be used when the ratio of the barcode label image to the target image is large and the height of one line of the barcode label is sufficient.

Further, when the ratio of the barcode label image to the target image is small and the height of one line of the barcode label is not sufficient, various methods can be considered. As one of them, first, two field screens (even component and odd component) are separately transferred and captured. In this case, only the odd components are loaded into the frame memory 6 without leaving any rows, and then only the even components are loaded without leaving any rows. The contents of the frame memory 6 at that time are as shown in FIG.
However, in FIG. 7, it does not matter which of the odd component and the even component is on top.

The reason why both the odd and even components are transferred here is that the two field images that originally image the same object are considered as different images, and if the image shown in FIG. 7 cannot be read,
This is because reading reliability can be improved by reading the image below.

In this case, if the first judgment in step S15 is (NO) in the flow chart of FIG. 4, the optimum scan of step S12 is tried again on the lower screen of FIG. If the speed of the moving barcode label is known, the position information A, B, C, D of the barcode label on the upper screen of FIG. 7 obtained in step S3 is known.
Since the positions A ′, B ′, C ′ and D ′ of the bar code label on the lower screen of FIG. 7 can be calculated by considering the moving speed and scanning speed of the bar code label,
There is no need to call the barcode label detection routine again.

If the moving speed of the bar code label is sufficiently negligible with respect to the scanning speed, the position information of the bar code label on the upper screen of FIG. 7 can be used as it is.

Also, one field image is taken into the frame memory 6 at intervals of one line, and the contents of the frame memory of the line immediately above is copied to the empty line,
There is also a way to increase the read reliability. In this way,
Since copying can be easily realized by hardware, the high speed reading of the bar code label is not lost.

If the height of one line of the bar code label is sufficient, as shown in FIG. 8, an arbitrary number of pixels are added to all pixel information to make one pixel, and the frame memory 6 There is a way to store. This can also be easily realized by hardware, and also has the effect of suppressing noise. FIG. 8 shows an example in which a total of 4 pixels in vertical 2 rows and horizontal 2 columns are added to form one pixel.

When the non-interlaced two-dimensional image pickup device 5 is used, step S21 is a frame image capture. Next, the barcode label detection routine called in step S3 will be described with reference to the flowchart shown in FIG. 9 and the barcode label projection image shown in FIG. This bar code label detection routine detects the presence or absence of a bar code label, and detects the position information of the bar code label, that is, image data in parallel to the bar code label.
It includes two types of bar code label detection processing, namely, an extraction range (variables STARTTOP and STOPBOTTOM) for extracting from the bar code and a slope (variable SLOPE) of the bar code label. Where the variable STARTTO
The value of P indicates the top coordinate of the barcode label, and the content of the variable STOPBOTTOM indicates the bottom coordinate of the barcode label. The content of the variable SLOPE indicates the inclination of the barcode label.

In this bar code label detection routine, first, a bar code label screen detection routine, which will be described later, is called to detect a bar code label from the screen (label image data) stored in the frame memory 6 (step). S31). It is judged whether or not the detected bar code label is completely within the screen (step S
32) If it is not in the screen (NO), the process returns with the information without the barcode label. However, if it is within the screen (YES), the scan and detection routine described later is called to detect whether or not the start code exists in the label image data of the frame memory 6 (step S33). That is, the coordinates e and g shown in FIG. 10 are detected. When the start code is detected and confirmed by this routine, the coordinate variables e and g on the frame memory 6 shown in FIG. 11 are defined. Here, the coordinate variable e is the coordinate of the smallest Y coordinate among the coordinates where the start code 22 is detected, and g is the coordinate of the largest Y coordinate among the coordinates where the start code 22 is detected. Are shown respectively.

Next, it is judged whether or not both the coordinate variables e and g are defined (step S34). If not defined (NO), the bar code label is absent and the process returns to the main flow chart. That is, the process returns with the information without the bar code label. As described above, the presence or absence of the barcode label is detected.

Next, the extraction range (variable STAR) for detecting the position information of the bar code label, that is, for extracting the image data from the frame memory 6 in parallel with the bar code label.
TTOP and STARTBOTTOM) and the slope of the barcode label (variable SLOPE) are calculated.

That is, in step S34, if it is determined that both the coordinate variables e and g are defined (YES), the position information of the bar code label is detected, that is, parallel to the bar code label. Extraction range for extracting image data from the frame memory 6 (variable STARTTOP
And STARTBOTTOM) (step S3
5) and the inclination of the barcode label (variable SLOPE) are calculated (step S36). Then, it returns with the information with the barcode label.

Next, regarding the bar code label in-screen detection routine called in step S31 in the above-mentioned bar code label detection routine and the judgment of whether or not the bar code label in the screen in step S32 is completely inserted, The detection in the barcode label screen shown in the flowchart of FIG. 12 and FIG. 13 will be described.

As described above, this bar code label in-screen detection routine confirms whether or not the moving bar code label is completely contained in the frame memory 6. Even if the bar code label is not completely stored in the frame memory 6, it is possible to estimate the position and read the bar code symbol information depending on the number of codes out of the screen by using the intelligent scan described later. Although it is possible, since the reading speed becomes slower by that amount, such a routine is provided to correspond to the moving bar code label. Here, it is assumed that the bar code label moves vertically with respect to the image.

First, in the top line of the image, the start code 22
Alternatively, the stop code 23 is detected (step S4
1). Details of this detection method will be described later. Next, it is determined whether or not it is detected (step S42). Here, if it is determined that either one is detected (YES), it means that the upper portion of the barcode label is not within the image as shown in FIG. Therefore, in such a case, the process returns with the information that the bar code label is not in the screen. If it is not detected (NO), the start code 22 or the stop code 23 is detected in the bottom line of the image (step S43).

Next, it is judged whether or not the detection is made (step S44). As mentioned above, if it is determined that either of them is detected here (YES), it means that the bottom of the barcode label is not in the image, and the information that the barcode label is not in the screen is displayed. To return.

If it is not detected (NO),
The start code 22 or the stop code 23 is detected in the center line of the image (step S45). Then, it is judged whether or not it has been detected (step S46), and if not detected here (NO), the process returns with the information that the bar code label is not in the screen. However, if detected (YES), the process returns with the information that the barcode label is within the screen.

Next, the scanning and detecting routine called in step S33 in the above-mentioned bar code label detecting routine will be described with reference to the flowchart shown in FIG. 14 and the drawing for explaining the scanning method shown in FIG.

This scanning and detecting routine detects both the coordinate variables e and g as described above. That is, first, a variable i used to count the number of loops and a variable sta that stores the number of detected start codes.
rt num is initialized (step S51). next,
The row number n to be scanned and the scan interval nn are determined (step S52). Row here num is the number of lines of the target image. As shown in FIG. 15, the line numbers scanned by this subroutine are of a system in which the search width is gradually narrowed. This is similar to the NEWTON method that iteratively solves equations.

By using such a method, the start code 22 can be detected at high speed regardless of the position of the bar code label image in the frame memory. Next, j is initialized in order to scan 2 i times for each loop i (step S5).
3). Then, it is determined whether j is "0" (step S
54). Here, if j is not 0, that is, if the proposition in step S54 is true (NO), the line number n is changed to step S.
A value obtained by doubling the scan interval nn obtained in 52 is added (step S55). When j is "0", that is, when the proposition in step S54 is false (YEA), step S55 is skipped, the process proceeds to step S56, and the image data of the nth row is fetched (step S56).

Next, the start code 22 is detected in the captured image data (step S57).
Next, it is determined whether or not the start code 22 is detected (step S58), and if it is not present (NO), the process proceeds to step S62. If the start code 22 exists (YES), the detected coordinates are set to the coordinate array sta.
rt It is stored in pos (step S59). And
The variable start Add 1 to num (step S
60).

Next, the variable start It is determined whether or not the number of num exceeds the predetermined number (step S61). Here, if it exceeds (YE
S), and then proceeds to step S66. However, if it has not exceeded (NO), the variable j is incremented (step S62), and it is determined whether or not j is smaller than the i-th power of 2 (step S63). If j is smaller (YES), the process returns to step S54. However, if j is not smaller than the i-th power of 2 (NO), the variable i is incremented (step S64), and it is determined whether i is smaller than the specified constant Loop (step S65). Here, if i is small (YE
S), and returns to step S52. However, if i is not smaller than the specified constant Loop (NO), the process returns with the information that the barcode label is not detected. Therefore, in this subroutine, maximum (2 to the power of Loop) -1
It will be scanned twice.

Further, in step S61,
rt If the number of num exceeds the specified number, a routine for obtaining the maximum and minimum values of the Y coordinate, which will be described later, is called (step S66). Then, it returns with the information of the barcode label detection. In this way, when the start code can be detected by the predetermined number or more, the processing can be speeded up by returning the control to the upper routine even during the loop processing.

Next, the start code detection routine called in step S57 in the scan and detection routine in FIG. 14 will be described with reference to the flowchart shown in FIG.

First, the captured image data is converted into width information (step S71). Next, the converted width information is normalized (step S72) and matching is performed (step S73). Then, it is determined whether or not the width information matches the start code (step S74),
If the start code is determined (YES),
Return with the information including the start code. If it is not the start code (NO), return with the information without the start code.

Next, the width information conversion routine called in step S71 will be described with reference to a series of flow charts shown in FIGS. This routine is to find the width between the bar and the space of the barcode label, and find the boundary between the bar and the space by the differential signal,
At this time, the peak of the data is obtained by approximating a quadratic curve. Then, the width is obtained by sequentially obtaining the peaks and obtaining the difference between the positions.

First, the fetch buffer array scan fetched by the upper routine The line is an array of line data, and the value fetched and stored in the variable num is defined as the number of data (step S81). Next, the result of decrementing the value of this variable num is stored in the position indicator counter i, and the variable j is initialized to "0" (step S82).

Then, the fetch buffer array scan
Capture buffer array scan from the value of the i-th position of the line (indicated by the position indicator counter i) l
The value at the i-1 position of ine is subtracted, the solution is fetched, and the buffer array scan The line is reset to the i-th position (step S83). Thereafter, the position indicator counter i is decremented (step S84), and the result i
Is determined to be greater than "0" (step S8).
5). If i is not larger than 0 (NO), the process returns to step S83, that is, the line data is first differentiated in steps S81 to S85. Also the result i
If is larger than "0" (YES), the position indicator counter i is initialized to 2 (step S86).

Then, the fetch buffer array scan
The value of the i-th position of the line is the threshold variable THRESH
It is larger than the value of OLD, and the capture buffer array s
can It is larger than the value at the (i-1) th position of the line, and the capture buffer array scan i of line
It is determined whether or not the value at the + 1st position is greater than or equal to the value (step S8).
7), if so (YES), set the code index flag to U
P (step S88). However, if not (NO), the process proceeds to the next step S89.

Here, the capture buffer array scan
The value at the i-th position of the line is smaller than the value of the threshold variable THRESHOLD (-THRESHOLD) with a negative sign, and the capture buffer array scan li
It is smaller than the value at the (i-1) th position of ne, and the capture buffer array scan It is determined whether or not the value at the (i + 1) th position of the line is less than or equal to the value (step S89), and if so (YES), the code index flag is set to DOWN (step S90). Otherwise (NO), the process moves to the next step S91.

Here, the position indicator counter i is incremented (step S91). It is determined whether i is smaller than the number obtained by subtracting 1 from the value of the variable num (step S9).
2). If it is smaller (YES), the process returns to step S87, and if it is not smaller (NO), the process returns to the routine of step 72.

That is, from step S87 to step S
At 92, the first peak that exceeds the value of the threshold variable THRESHOLD, that is, the first peak in FIG. 19, is detected.
Then, when the sign of the detected peak is positive, the code index flag is set to UP, and the process proceeds to the next step S93. When the sign of the detected peak is negative, the code index f
lag is set to DOWN, and the process proceeds to step S94. If no peak is detected even after scanning the line data, the control is returned to the upper routine.

If the code index flag is set in this way, then the value of the position indicator counter i is set to x1, and the fetch buffer array scan is set. i of line
-The value at the 1st position is y1, the capture buffer array sc
an The value at the i-th position of the line is y2, the capture buffer array scan By setting the value at the (i + 1) th position of the line to y3, the detected peak position and the data on both sides of the detected peak position are fitted with a quadratic curve (step S93). And -0.5 (-y1-2x1y1 +
4x1y2 + y3-2x1y3) / y1-2y2 + y3
Is calculated to obtain the peak position of the quadratic curve,
It is stored in the variable lastpos (step S9).
4).

After that, the value of the position indicator counter i is changed to the variable n.
It is smaller than the number obtained by subtracting 2 from the value of um, and the variable j
Is WIDTH It is determined whether or not it is smaller than NUM (step S95), and if the variable j is not small (NO),
Return to the upper routine.

Here, the variable j is WIDTH Explain the proposition that it is smaller than NUM. WIDTH NUM
Is a constant in which the number of widths to be obtained is stored, and j is a variable in which the number of obtained peaks is stored. here,
In order to detect the position of the start code 22, there is a condition that the start code 22 is always located on the left side of the frame memory 6 and that there is no information other than the bar code label in the frame memory 6. , The sequence scan It is not necessary to convert all the differential information stored in line to width information.

That is, since the minimum number of peaks required to detect eight pieces of width information is nine, the constant WIDTH
It means that the number of NUM may be more than that.
Usually WIDTH NUM is determined by the SN ratio of the image. That is, if there is no information other than the barcode label in the frame memory 6, WIDTH NUM is "9"
Can be set to. Otherwise, WIDTH NUM must be a number greater than or equal to "9".

For the above reason, the above proposition is included in the judgment of step S95. By entering this judgment into the proposition, the routine can be ended without obtaining all the width information, so that the processing speed can be increased. As will be described later, if you want to obtain all the width information from the differential information, you can of course remove this proposition.

If the proposition of step S95 holds (YES), the position marker counter i is incremented (step S96). Then, the code index flag is DOWN, and also the capture buffer array scal
n The value of the i-th position of the line is the threshold variable THRE.
Greater than the value of SHOLD and the capture buffer array scan It is larger than the value at the (i-1) th position of the line and is the acquisition buffer array scan i + of line
It is determined whether or not the value at the first position is greater than or equal to the value (step S97). If all the terms are satisfied (YES), the code index flag is set to UP (step S98).

However, if all the terms are not satisfied in step S98 (NO), then the code index flag is UP and the fetch buffer array scan is also present.
The value of the i-th position of the line is smaller than the value of the threshold variable THRESHOLD with a negative sign and the fetch buffer array scan It is smaller than the value at the (i-1) th position of the line and is the acquisition buffer array scan li
It is determined whether or not it is less than or equal to the value of the i + 1th position of ne (step S99). If all the terms are not satisfied (NO), the process returns to step S95, and if all the terms are satisfied (YES), the code index flag is set to DOWN (step S100).

When the code index flag is reset in this way, next, the value of the position indicator counter i is set to x1,
Also capture buffer array scan line i-1
The value at the second position is y1, the capture buffer array scan
The value at the i-th position of the line is y2, the capture buffer array scan The value at the (i + 1) th position of the line is y
By setting 3, the detected peak position and the data on both sides of the detected peak position are fitted with a quadratic curve (step S
101).

Then, -0.5 (-y1-2x1y1 +
4x1y2 + y3-2x1y3) / y1-2y2 + y3
Is calculated to obtain the peak position of the quadratic curve,
It is stored in the variable nowpos (step S10).
2).

As a result, as shown in FIG. 20, the peak position lastpos and the peak position n obtained this time are determined.
owpos is obtained. Then, the peak-to-peak distance is obtained by taking the difference between the two peak positions thus obtained, and stored in the position indicated by the variable j of the width information storage array variable width (step S103).

Then, the peak position variable lastpos is updated to the value of the peak position variable nowpos (step S
104) and after incrementing the variable j (step S105), the process proceeds to step S95.

In this way, in steps S95 to S105, the peaks are sequentially detected, and the peak-to-peak distance is stored in the width information storage array variable width. Here, the point different from the first peak detection in steps S87 to S92 is, for example, the current code index flag.
Is DOWN, the peak that must be found next is the point with a positive sign.

Next, the normalization routine called in step S82 in the start code detection routine as described above will be described with reference to the flowchart shown in FIG. First, "0" is set in the loop count variable i and the variable sum.
Is substituted and initialized (step S111). Next, the array width [i] in which the width information obtained by the conversion routine is stored is added to the width sum in the variable sum, and the result is stored again in sum (step S112).

Next, the variable i is incremented (step S113). Next, it is determined whether or not this variable i is smaller than "8" (step S114), and if this variable i is smaller than "8" (YES), the above step S
Return to 112. If the variable i is larger than “8” (N
O), and proceeds to the next step S115. The variable i is changing from "0" to "7" in the PD.
This is because the code word of F417 is composed of a total of eight pieces of width information of four bars and four spaces, and the start code is normalized using all eight pieces of width information.

Then, again, "0" is assigned to the variable i for initialization (step S115). Next, "17" is divided by the variable sum to obtain the i-th wi of the width information array.
The value obtained by multiplying the value stored in dth [i] is rounded to the nearest integer (step S116). Here, round in FIG. 21 means rounding. Then, the variable i is incremented (step S117),
It is determined whether the variable i is smaller than "8" (step S118). If this variable i is smaller than "8" (YE
S), the process returns to step S116, and the variable i is "8".
If it is larger (NO), the process returns to the upper routine.

Next, the start code matching of step S73 in the start code detection routine shown in FIG. 16 as described above will be described. Although any matching method may be used, the bar portion often becomes thicker than intended when printing a barcode label. If the bar portion is too thick, the decoding performance will be significantly reduced. However, in such a case, the space becomes thinner as the bar gets thicker, so
It is better to adopt a method in which the width information obtained by adding the bar thickness and the space thickness is used as a matching target.

Next, the routine for detecting the maximum and minimum values of the Y coordinate called in step S66 in the above-mentioned detection routine of FIG. 14 will be described with reference to the flow chart shown in FIG. This Y-coordinate maximum / minimum value detection routine is carried out in the coordinates of the start code detected by the specified number,
This is a routine for obtaining the maximum and minimum Y coordinates.

First, in the Y coordinate values (MAXy and MINy) of the coordinate variables MAX and MIN, 0 of the array of detected coordinates of the start code detected in step S57 (FIG. 14) is set.
Start stored in th Value of Y coordinate of pos [0] (start Substitute pos [0] y) (step S121). Next, 1 is substituted into the variable i required for looping by the number of data (step S122).

Next, the start stored in the i-th position of the array Y coordinate value of pos [i] (start
It is determined whether pos [i] y) is larger than MAXy (step S123). Here, when it is larger than MAXy (YES), that is, when the proposition of step S123 is true, the i-th start of the coordinate array is set. pos
Substitute [i] into the coordinate variable MAX (step S12
4) and the process proceeds to step S127. However, MAXy
If smaller (NO), that is, if the proposition in step S123 is false, the process proceeds to step S125.

Next, the start stored in the i-th position of the array Y coordinate value of pos [i] (start
It is determined whether pos [i] y) is smaller than MINy (step S125). If smaller than MINy (YES), that is, if the proposition of step S125 is true, the i-th start of the coordinate array is set. pos
Substituting [i] into the coordinate variable MIN (step S12
6) and proceeds to step S127. However, when it is larger than MINy (NO), that is, when the proposition in step S125 is false, the process proceeds to step S127.

Next, the variable i is incremented (step S127). Next, the variable i is the value in step S
Start obtained from 57 It is determined whether or not it is larger than num (step S128), and if the variable i is smaller (NO), that is, if the proposition of step S128 is false, the process proceeds to step S123. However, if the variable i is larger (YES), that is, if the proposition of step S128 is true, after storing the coordinate variable MAX in the coordinate g and the coordinate variable MIN in the coordinate e in FIG. Step S129), and the process returns to the upper routine of FIG.

Next, a routine for obtaining the label range called in step S35 in the label detecting routine shown in FIG. 9 will be described with reference to the flowchart shown in FIG.

First, the equation of a straight line parallel to the line segment e-g is defined, for example, the equation y = ax + b of the line segment e-g is obtained (step S131). Next, the intercept b is defined so that this straight line crosses the start big bar 22A (step S132). The structure of the start code 22 is
For example, start big bar 22A consisting of 8 bars
And 3 pairs of white and black bars, 3 white bars in total 17
It is assumed that each bar is composed of a number of bars, and the result of imaging this is N pixels. It is known that the straight line represented by the equation y = ax + b moves in parallel by changing the intercept b. Therefore, in order to obtain a straight line that crosses the big bar 22A, the line segment eg is set to {(1
7-8 / 2) / 17} × N pixels may be moved to the left by an intercept b.

In this way, the start big bar 22
Once a straight line crossing A is obtained, then the intersections of the straight line and the equations defining the screen are A, A ′, respectively.
(See FIG. 10) (step S133).

Then, the data is sequentially examined from the midpoint of the line AA 'toward the point A (step S13).
4) Then, it is checked whether or not there is an edge (step S135). This check can be performed by, for example, an intensity comparison for observing a change in luminance, a differential method, a second differential method, or the like. In this way, when the edge is detected, the detected coordinates are stored in the coordinate variable i (step S13).
6). That is, the detected coordinate is set to point i.

Next, from the midpoint of the line AA ', the data is quasi-directed to A' (step S13).
7) Then, it is checked whether or not an edge exists (step S138). Thus, if an edge is detected,
The detected coordinates are stored in the coordinate variable m (step S13).
9). That is, the detected coordinate is set to the point m.

Then, the point e indicated by the coordinate variables e and g,
A perpendicular line is drawn from the point i indicated by the coordinate variable i on a straight line passing through g, and the coordinates of the intersection are stored in the coordinate variable a (step S140). That is, an equation of a straight line orthogonal to the line A-A 'passing through the point i is obtained, an intersection of the straight line passing through the points e and g is obtained, and the intersection is defined as a point a.

Similarly, points e and g indicated by coordinate variables e and g
A perpendicular line is drawn from the point m indicated by the coordinate variable m on a straight line passing through g, and the coordinates of the intersection are stored in the coordinate variable d (step S141). That is, the equation of a straight line orthogonal to the line A-A 'passing through the point m is obtained, the intersection of the straight line passing through the points e and g is obtained, and the intersection is defined as the point d.

After storing the value of the coordinate variable a thus obtained in the coordinate variable STARTTOP and the value of the coordinate variable d in the coordinate variable STARTBOTTOM (step S142), the control is returned to the upper routine.

Next, a routine for obtaining the label inclination called in step S36 in the above-described label detection routine will be described with reference to the flowchart shown in FIG. 24 and FIG.
It will be described with reference to the explanatory diagram for obtaining the inclination of the label shown in FIG.

First, the value of the x coordinate of the coordinate variable STARTBOTTOM is divided by the value obtained by subtracting the value of the y coordinate of the coordinate variable STARTTOP from the value of the y coordinate of the coordinate variable STARTBOTTOM, and the coordinate variable STARTTO
The value of the x coordinate of P is divided by the value of the result of subtracting the value of the y coordinate of the coordinate variable STARTTOP from the value of the y coordinate of the coordinate variable STARTBOTTOM, and the difference between these two quotients is stored in the slope variable Slope (step S151). ).

Next, the result of multiplying the y coordinate of the coordinate variable STARTBOTTOM and the x coordinate of the coordinate variable STARTTOP is a value obtained by subtracting the value of the y coordinate of the coordinate variable STARTTOP from the value of the y coordinate of the coordinate variable STARTBOTTOM. Split and coordinate variable STARTBOTTO
The result of multiplying the x coordinate of M by the y coordinate of the coordinate variable STARTTOP is divided by the value of the result of subtracting the value of the y coordinate of the coordinate variable STARTTOP from the value of the y coordinate of the coordinate variable STARTBOTTOM, and the difference between these two quotients. Is stored in the intercept variable intersect (step S152).
The superscript of asterisk * in the figure means the multiplication symbol x.

Next, the threshold value determination routine called in step S5 of FIG. 4 will be described with reference to the flowchart shown in FIG. 25 and the explanatory diagram for calculating the threshold value shown in FIG.

First, a value corresponding to the point e shown in FIG. 10 is substituted for the point p (step S161). Next, coordinate variable p
The line indicated by the y-coordinate value of, that is, the data of the data-acquisition line as shown in FIG. Yes (step S163).

Then, the data from the point p to the point q is differentiated (step S164). As a result, for example, in FIG.
In the case of the data acquisition line shown in FIG.
A differential waveform of the start code 22 as shown in 6 (b) is obtained. Then, the absolute value of the third peak of this differential data is set as the variable MAX (step S165).

Here, the reason for looking at the third differential data is that, conceptually, the threshold is to be decided at the point where the contrast becomes the lowest in the barcode area (that is, the edge differential peak is the lowest). This is because the third edge of the start or stop code, which has the smallest space between the label bar and the space, is selected. This enables stable decoding regardless of label size and label illumination conditions.

Then, the value of the variable MAX thus obtained is set to a constant THRESHOLD of the ratio to the peak. RAT
It is divided by IO and the result is substituted for the threshold variable THRESHOLD (step S166). That is, the threshold is tentatively obtained from the obtained data. In addition, the variable THRESHOL
D RATIO is a value indicating which fraction of the peak is selected as the threshold value, and is usually set to "2" or "3".

Next, the tentatively obtained threshold value THRESHOLD
Is the minimum threshold constant THRESHOLD SMALL
Larger than the maximum threshold value (step S167) and the maximum threshold constant THRESHOLD It is determined whether it is smaller than BIG (step S168). That is, it is determined whether or not the tentatively obtained threshold value is within the range of the threshold value. Constant THERESHOLD If the maximum value indicated by SMALL is exceeded (YES), the constant THRESH
OLD The maximum value indicated by SMALL is the threshold variable THE
Substitute for RESHOLD, that is, set the threshold to the minimum value (step S169).

In addition, the variable THRESHOLD If less than the minimum value indicated by BIG (YES), THERES
HOLD The minimum value indicated by BIG is the threshold variable THRE.
Substitute in SHOLD, that is, set the threshold value to the maximum value (step S170).

Next, the indicator information determination routine called in step S6 of FIG. 4 will be described with reference to FIG.
It will be described with reference to the row chart shown in FIG.

First, as the starting point of the reference coordinates for reading the row indicator information, the top coordinates TOP of the label
Is stored in the coordinate variable WORK (step S17).
1). Next, a straight line I passing through this coordinate variable WORK and having a slope indicated by the slope variable Slope of the label is defined (step S172), and points W1 and W2 at which this straight line I crosses the screen frame are defined (step S173). ).

Then, the image data on the line segment W1-W2 is fetched (step S174), and the row indicator information contained therein is read (step S175). Here, the reading of the row indicator information is performed as follows, for example. That is, an edge is detected from the image data on the target line captured in step S174, that is, a black and white pixel value, and converted into width information. Then, the start code 22 is detected from this width information, and since it is known that the code next to the start code 22 is the row indicator 21A, it is read.

Similarly, the stop code 23 is detected, and it is also known that the code immediately before the stop code 23 is the row indicator 21A, so that it is read. In this way, when the row indicator 21A is read, it is compared with a bar code table (not shown), and the matching portion is converted into a code, that is, information such as the number of rows, the number of columns, and the security level. There are various methods for conversion into width information, but the conversion can be performed by calling a width information conversion routine as described later.

Next, it is checked whether or not the row indicator information has been established (step S176), and if it has been confirmed (YES), the process proceeds to step S180. If not confirmed (NO), the process proceeds to step S177. Here, “determined” means that the row indicator 21A is read several times and the reliability of the information is sufficiently improved. For example, if the information (the number of rows, the number of columns, the security level) written in the row indicator 21A is read 10 times, and if the same information is obtained 10 times,
Suppose it has been confirmed.

If the row indicator information is not confirmed in the determination in step S176 (NO), the y coordinate value of the coordinate variable WORK is incremented by a predetermined increment L. INC is added, and the result is substituted as the y coordinate value of the new coordinate variable WORK (step S177). Further, the value of the label inclination variable Slope is multiplied by the value of the y coordinate of the coordinate variable WORK, and the result is the label intercept variable intersect.
Is added and the result is substituted as the x coordinate value of the new coordinate variable WORK (step S178). In this way, the reference coordinate for newly scanning is reset to the coordinate variable WORK.

Then, this reset coordinate variable WOR
It is determined whether the y coordinate value of K exceeds the y coordinate value of the bottom coordinate variable BOTTOM of the label, that is, whether it is within the label area (step S179), and if it is within the label area, repeat from step S172, If it is outside the label area, it returns to the upper routine with undefined matrix information.

On the other hand, if it is determined in step S176 that the row indicator 21A has been confirmed (YE
S), the number of recoverable data is calculated from the security level obtained from the raw indicator information, and the REST
It is stored in NUM (step S180). Further, the number of rows of the label obtained from the row indicator information is the row number variable ROW of the label. It is stored in NUMBER (step S181). Further, the number of label columns is extracted from the row indicator information, and the label column number variable C
OLUMN Stored in NUMBER (step S1)
82). After that, control is returned to the upper routine with the information that the matrix is defined.

Next, the inclination redefinition routine called in step S8 of FIG. 4 will be described with reference to the flow chart shown in FIG. 29 and the explanatory diagram for redefining the inclination shown in FIG. Here, the reason for redefining the slope will be described below.

The inclination obtained from e and g in FIG. 10, that is, the inclination obtained only from the information of the start code 22 is not an accurate value due to problems such as print quality and CCD accuracy.
Therefore, if the indicator information is determined in step S8, the stop code 23 is detected from a position close to the stop code 23, and the inclination is obtained from the information of both the start code 22 and the stop code 23, a more accurate inclination can be obtained. The value is obtained. Obtaining a more accurate tilt value before the optimum scan in step SC leads to faster scan.

First, the detection start point p is determined (step S191). p is a straight line I connecting the left upper end point u and the lower end point v of the column of the right row indicator, which is obtained by the above-mentioned indicator information determination routine, and is divided into four equal parts along the line I from u. This is the point where Then p,
The data between p'is fetched (step S192). p '
Is a point which extends from the point p in parallel with the x-axis and intersects a straight line of y = column number.

Next, a stop code detection routine, which will be described later, is called (step S193). Then, it is judged whether or not the stop code is detected (step S19
4) If it is not detected (NO), return with the information that the inclination is not redefined. If detected (YES), the coordinates of the detection point are stored in f shown in FIG.
f is defined (step S195).

Next, the detection start point q is determined (step S196). This point q is divided into four equal parts of the straight line I connecting the upper end point u of the left end and the lower end point v of the column of the right row indicator obtained by the above-mentioned indicator information determination routine, and divided into four equal parts along v from I. It is a point that has advanced the distance. Next, the data between the points q and q'is fetched (step S197).
A point q ′ is a point which extends from the point q in parallel with the x-axis and intersects a straight line of y = column number. Next, the stop code detection routine in step S193 is called to detect the stop code (step S198). Then, it is judged whether or not the stop code is detected (step S199), and when not detected (NO), the process returns with the information that the inclination is not redefined. If detected (YES), the coordinates of the detection point are stored in h shown in FIG. 10 and h is defined (step S200).

Next, a label range detection routine, which will be described in detail later, is called (step S201), and the inclination is calculated (step S202). Then, it returns with the information of the tilt redefinition.

The stop code detection routine called in steps S193 and S198 will be described with reference to the flow chart shown in FIG. This routine is basically the same as the start code detection shown in FIG. 16, but the matching part is different from that in FIG. This is because the start code 22 and the step code 23 of the PDF 417 have different width information. As for the matching method, it is desirable that the matching method is not affected by the printing accuracy of the barcode label as described above.

Next, the label range detection routine called in step S201 of FIG. 29 will be described with reference to FIG.
This will be described with reference to the flowchart shown in FIG. First, the equation of a straight line parallel to the line segment f-h is defined, for example, the line segment f-
An equation y = ax + b of h is obtained (step S22
1).

Next, this straight line is the stop big bar 23.
The intercept b is defined so as to cross A (step S
222). The structure of the stop code 23 is composed of, for example, a stop big bar 23A composed of eight bars, three pairs of white bars and black bars, and three white bars, for a total of 17 bars. It is assumed that the imaged result is N pixels. It is known that the straight line represented by the equation y = ax + b moves in parallel by changing the intercept b.

Therefore, in order to obtain a straight line that crosses the big bar 23A, the line segment f-h is set to {(17-8 /
2) / 17} × N pixels may be moved to the left by an intercept b.

When a straight line crossing the stop big bar 23A is obtained in this way, next, the intersections of the straight line and the equations defining the screen are taken as B and B '(see FIG. 10) (step S223). .

Then, the data is sequentially examined from the midpoint of this line BB 'toward the point B (step S22).
4) Then, it is checked whether or not there is an edge (step S225). This check can be performed by, for example, an intensity comparison for observing a change in luminance, a differential method, a second differential method, or the like. In this way, when the edge is detected, the detected coordinates are stored in the coordinate variable j (step S22).
6). That is, the detected coordinate is set to point j.

Next, the data is sequentially examined from the midpoint of the line BB 'to the point B' (step S2).
27), it is checked whether or not an edge exists (step S228). In this way, when the edge is detected, the detected coordinate is stored in the coordinate variable k (step S
229). That is, the detected coordinate is set to the point k.

Then, the point f indicated by the coordinate variables f and h,
A perpendicular line is drawn from the point j indicated by the coordinate variable j on a straight line passing through h, and the coordinates of the intersection are stored in the coordinate variable b (step S230). That is, the equation of a straight line orthogonal to the line BB ′ passing through the point j is obtained, the intersection of the straight line passing through the points f and h is obtained, and the intersection is defined as the point b.

Similarly, the points f, which are indicated by the coordinate variables f, h,
A perpendicular line is drawn from the point k indicated by the coordinate variable k on a straight line passing through h, and the coordinates of the intersection are stored in the coordinate variable c (step S231). That is, the equation of a straight line orthogonal to the line BB ′ passing through the point k is obtained, the intersection of the line passing through the points f and h is obtained, and the intersection is defined as the point c.

After storing the value of the coordinate variable b thus obtained in the coordinate variable STOPTOP and the value of the coordinate variable c in the coordinate variable STOPBOTTOM (step S232), the process returns to the upper routine of FIG.

Next, the label inclination recalculation routine called in step S202 of FIG. 29 will be described with reference to FIG.
This will be described with reference to the flowchart shown in FIG. First, the value of the y coordinate of the coordinate variable STOPTOP is set to the coordinate variable STOP.
The value of the x coordinate of the coordinate variable STARTTOP is subtracted from the value of the x coordinate of PTOP, and the result is divided, and the value of the y coordinate of the coordinate variable STARTTOP is divided by the coordinate variable STOPT.
It is divided by the value of the result of subtracting the value of the x coordinate of the coordinate variable STARTTOP from the value of the x coordinate of OP, and the difference between these two quotients is stored in the slope variable Slope1 (step S24).
1).

Next, y of the coordinate variable STOPBOTTOM
The coordinate value is divided by the result of subtracting the x coordinate value of the coordinate variable STARTBOTTOM from the x coordinate value of the coordinate variable STOPBOTTOM, and the coordinate variable STARTB
The y coordinate value of OTTOM is the coordinate variable STOPBOTT
From the value of the x coordinate of OM, the coordinate variable STARTBOTTOM
The value of the x-coordinate is subtracted, and the difference between these two quotients is stored in the slope variable Slope2 (step S242).

Next, it is determined whether the absolute value of the difference between the slope variable Slope1 and the slope variable Slope obtained in step S36 is smaller than the threshold constant SLOPETHRESH1, and the absolute value of the difference between the slope variables Slope2 and Slope is smaller than the threshold constant SLOPETHRESH1. It is determined (step S243). Here, abs in the figure means an absolute value. When the proposition in step S243 is false (NO), the information is returned without recalculation. Otherwise (YES), the variable Slope
The absolute value of the difference between 1 and Slope2 is taken, and it is determined whether or not it is larger than the threshold constant SLOPETHRESH2 (step S244).

If the absolute value is smaller in step S244 (NO), the process proceeds to step S250, and if the absolute value is larger, the process proceeds to next step S245. next,
The absolute value of the difference between the slope variable Slope calculated in step S36 and the variable Slope1 is stored in the variable a (step S245). Next, the absolute value of the difference between the slope variable Slope obtained in step S36 and the variable Slope2 is stored in the variable b (step S246).

Then, it is judged whether or not the variable b is larger than the variable a (step S247). When the variable b is large (YES), Slope1 is substituted for the slope variable Slope (step S248). If the variable b is small (NO), the slope variable Slope
Subset2 is substituted for (step S249). Also, in step S244, the slope variables Slope and Sl are
When the absolute value of the difference in ope2 is smaller than the threshold constant (NO), the slope variable Slope is set to the slope variable Slope.
The sum of e1 and Slope2 divided by 2, that is, Slo
Substitute the average of pe1 and Slope2 (step S2
50). And it returns with the information of recalculation.

In the series of steps S244 to S249, as shown in FIG. 34, the start code 22 or the stop code 23 of the PDF417 bar code label is displayed.
Considering the case where there is a deficiency in one of the two and the range of the label is not obtained correctly. In the case of FIG. 34, the values of the slope Slope1 of the straight line n and the slope Slope2 of the n'are significantly different. This is detected by this routine, and FIG.
A value close to the slope Slope of the label calculated from the inside points e and g is adopted as a new slope. By using such a method, even if there is a defect in either the start code 22 or the stop code 23, a more accurate label inclination can be obtained.

The case where the determination in step S243 is false means that there is a defect in the upper and lower sides of either the start code 22 or the stop code 23. In this case, the inclination is not recalculated. .

Next, the scan equation determination routine called in step S9 of FIG. 4 will be described with reference to the flow chart shown in FIG. 36 and the image projected on the frame memory shown in FIG. Note that FIG. 35 shows an example in which the start code 22 is selected based on the line scan.

That is, first, the variable counter is initialized to "0" (step S251), and the reference coordinate variable WO
RK is initialized to the value of the top coordinate variable TOP of the label (step S252).

Next, the bottom coordinate variable BOTTO of the label
The difference between the y coordinate value of M and the y coordinate value of the top coordinate variable TOP of the label is the variable count. Substitute for end (step S253). That is, the number of patterns to be determined (the number of pixels in the label column direction in the case of FIG. 36) is calculated, and the variable count Store in end. That is, the number of patterns is a number for scanning the entire label surface.

Next, y of the top coordinate variable TOP of the label
The value of the variable counter is added to the coordinate value, and the result is substituted as the y coordinate value of the reference coordinate variable WORK (step S254). Also, the value of the label tilt variable Slope is multiplied by the value of the y coordinate of the reference coordinate variable WORK, the label intercept variable intersect is added to the result, and the result is substituted as the x coordinate value of the new reference coordinate variable WORK. (Step S255). Thus, the variable co
The reference coordinate variable WORK is reset with an increase in the unter value.

Next, this reset coordinate variable WORK
A straight line I having a slope indicated by the slope variable Slope of the label is defined (step S256), and this straight line I
Finds two points that intersect with the screen frame POINT P and DIM POINT Q
(The array number indicated by the value of the variable counter) is stored (step S257).

Then, the variable counter value is incremented.
(Step S258) and the result is reset.
Number of variable counter values required, that is, variable count
er It is checked whether or not end has been reached (step S
259). If the required number is not reached (YES),
Return to step S254, and if the required number is reached (N
O), the process proceeds to step S260. By the above operation
Combination of start and end points when scanning labels sequentially
The case is defined.

Next, the variable counter is initialized to "0" again (step S260), and the value of the label slope variable Slope is multiplied by the variable counter value to calculate the position of the label slope Slope calculated. Is calculated, and the result of this calculation is the array LINE of variables. Predetermined position of INC (variable counter
(S261).

Then, after incrementing the variable counter (step S262), the reset variable counter is the maximum size constant MAX of the array. NU
It is checked whether M has been reached (step S26).
3). If not reached (YES), the step S
Returning to step 261, and when reaching (NO), the routine returns to the routine of FIG.

From step S260 to step S described above.
By the loop of 263, an inclination pattern when one line of label information is taken in is obtained. The maximum size constant MAX of the array NUM indicates the size of a variable determined when the program is created. For example, assuming that the size of the frame memory 6 is “640 × 480 pixels”, “NUM” may be set.

Next, the embedding check routine called in step S10 of FIG. 4 will be described with reference to the flow chart shown in FIG. 37 and the image projected onto the frame memory shown in FIG. Note that FIG. 38 shows an example in which the start code 22 is selected as a reference in the line scan. The label size is “4 × 3”.

First, when a ROW detection routine, which will be described later, is called to scan the label, various information (detection line number, code length, row number, detection position, etc.) of the row found at the beginning and end is detected ( Step S271),
Variable ROW FIRST, ROW Store in LAST.

Next, the variable ROW FIRST and ROW
It is determined whether or not LAST is also confirmed (step S27).
2) If determined (YES), move to step S273. However, if it is not decided (NO), it is determined that decoding is not possible, and the process returns to the upper routine of FIG.

Thus, in step S272, the variable R
OW FIRST and ROW If LAST is also determined (YES), the head position coordinates of each code in the codeword matrix are estimated, and the two-dimensional array variable CODE
It is stored in the row number and column number th of each code of the POS (step S273). Here, as shown in FIG. 38, the estimation of the head coordinate position of each codeword is performed by the inclination of the label, the detection line number of the row found at the beginning and the end, the code length, the row number, the detection position, and the matrix size. (ROW NUMBER × COLUMN N
UMBER) and so on.

Next, among the coordinates estimated in the above step, the number of those that may cause the code to spill out of the screen is totaled, and the variable error is stored. It is stored in num (step S274). In the example shown in FIG. 38, (row number,
Column number) = (0,2), (1,2), (2,2),
Five codes of (3, 1) and (3, 2) are determined to be unreadable, and error 5 is stored in num.

Further, the error num is the variable REST of the repairable code number obtained in step S180 of FIG. Compare with NUM (step S275), er
ror num is REST If NUM or less (YE
S), decoding is possible and the process returns to the upper routine of FIG. On the other hand, error num is REST NUM
If it exceeds (NO), it is determined that decoding is not possible, and the process returns to the upper routine of FIG.

With such a routine, the position of each code can be estimated, and the number of codes estimated to be outside the screen can be counted. Therefore, it is possible to determine at this stage whether decoding is possible. This drastically reduces the worst pattern of not being decoded despite having been processed for a long time.

Next, the ROW detection routine called in step S271 of FIG. 37 will be described with reference to the flowchart of FIG. 39 and the diagram of the projected image on the frame memory shown in FIG.

First, the bottom coordinate variable BOTTO of the label
The difference between the y coordinate value of M and the y coordinate value of the top coordinate variable TOP of the label, that is, the number of pixels in the y direction of the label is calculated, and the calculation result is the variable pos. It is stored in num (step S281).

Next, the value of the variable inc is set to "1" (step S282), the value of the variable counter is set to "0" (step S283), and the value of the variable end is set to the step S2.
82 or pos obtained in S283 num (step S284).

Then, from the steps S282 to S28.
Each variable (initial value counter, increment in
c, end position end), a ROW number detection routine described later is called (step S285). If you can correctly detect the row number from the indicator information,
Variable ROW Information such as the detection line number, code length, row number, and detection position is stored in the POS.

That is, it is determined whether or not the row number can be correctly detected from the indicator information in the above routine (step S286), and if detected (YES), the process proceeds to step S287, and if not detected ( N
O), the decoding is not possible and the process returns to the upper routine.

Next, the data ROW obtained in the above step S285. POS to ROW It is stored in FIRST (step S287). That is, in steps S282 to S287, scan lines are sequentially set from the TOP of the label toward the BOTTOM, and the first detected indicator position is ROW. Stored in FIRST.

Further, from S288 to S293,
The scan line is set one after another from BOTTOM to TOP, and the first detected indicator position is ROW. Stored in LAST.

By such a routine, it is possible to know whether or not a part of the row cannot be processed (for example, it is projected outside the screen). Because the number of rows is 4
With the label ROW Row number 0 is detected by FIRST, and ROW If row number 2 is detected in LAST,
It can be seen that row number 3 cannot be processed because it cannot be detected.

Next, the ROW number detection routine called in step S285 in FIG. 39 will be described with reference to the flowchart shown in FIG. First, the scanning start point and the end point in the current variable counter value are set to the coordinate variable array DIM. POINT P, DIM PO
INT The increment in the x direction is "1" and the increment in the y direction is obtained from Q. With INC, one line of image data is taken out from the frame memory 6, and the fetch buffer array scan It is stored in the line, and the number of the data is stored in the variable num (step S301).

Then, a conversion routine for width information, which will be described later, is called (step S302) to convert the fetched data into width information. Next, based on this width information, the part that matches the bar code table (not shown) is converted into a code, and the indicator information is extracted from that information and saved (step S303). Here, it is determined whether or not the indicator information can be read (step S304), and if it cannot be read (NO), the variable counter is incremented,
That is, it is reset with the value of the interval variable inc (step S3
05), it is determined whether or not the reset variable counter value exceeds the variable end, that is, the label range (step S306). If not exceeded (NO), the process returns to step S301, and if exceeded (YES), it is determined that decoding is impossible, and the process returns to the upper routine of FIG.

On the other hand, if it is determined in step S304 that the data can be read (YES), the detection line number, code length, row number, and detection position at that time are set as variables ROW.
It is stored in the POS (step S307), the decoding is possible, and the process returns to the upper routine of FIG. Therefore,
ROW the various information of the indicator detected first after entering this routine It will be stored in the POS.

Next, the optimum scan routine called in step S13 of FIG. 4 will be described with reference to the flow chart shown in FIG. 41 and the image projected on the frame memory shown in FIG.

First, the indicator information variable R of the label
OW LAST detection position y coordinate value and ROW FIRS
The difference between the detected position y coordinate value of T, that is, the number of pixels in the y direction of the label is calculated, and the calculation result is stored in the variable end (step S311).

Next, the value of this variable end is set to the number of label lines (variable ROW). LAST row number-variable ROW FI
42 by dividing by the absolute value of the RST row number.
As shown in, the interval for scanning the center of each row only once is calculated and stored in the variable inc (step S3
12). And the variable ROW The detection line number of FIRST is set as the initial value of the variable counter (step S
313).

Next, in the current variable counter value
An array DIM of the coordinate variables for the scanning start point and the end point PO
INT P, DIM POINT Obtained from Q, x direction
The increment is "1" and the y-direction increment is the gradient increment array LIN.
E With INC, image data from frame memory 6
In take out and take it in buffer array scan l
It is stored in ine, and the number of data is stored in the variable num.
It is stored (step S314).

Then, the same as in step S302 described later, the conversion routine is called for the width information (step S302).
315), The extracted data is converted into width information.
Next, based on this width information, the part that matches the bar code table (not shown) is converted into a code and the information is saved (step S316).

Thereafter, the variable counter is incremented, that is, reset with the value of the interval variable inc (step S31).
7) It is determined whether or not the reset variable counter value exceeds the variable end, that is, the label range (step S318). If this judgment does not exceed (Y
ES), the process returns to step S314, and if it exceeds (NO), it is determined whether or not the information described in the label of the stored code information is completely decodable (step S318). If it is decodable (YES), it returns to the higher-level routine of FIG. 4 with the decodable information.
However, if the decoding is not possible (NO), the process returns to the upper routine of FIG. 4 with the information that the decoding is not possible.

Next, the intelligent scan routine called in step S14 of FIG. 4 will be described with reference to the flow chart shown in FIG. This skip scan routine is a routine for scanning only the code positions missed in the optimum scan.

First, of the codewords that could not be embedded in the optimum scan, the number of codes considered to be in the screen is defined as i (step S321). In addition, the row number and column number of each code are assigned to arrays R and C, respectively.
(Step S322). From the above, the number of codewords that could not be embedded in the optimum scan and their positions are obtained. If the column number and the row number are known, this position is set in the step S
CODE obtained in 10 The coordinates can be known from the POS.

Next, the variable counter and the inclination pattern counter n are initialized to "0" (step S32).
3). Then, the coordinate CODE of the counterth undetected place POS [[R [counter]] [C [c
a straight line I having a slope of Slope + dt [n] is defined (step S324). Points W3 and W4 at which this straight line I crosses the screen frame are defined, and the image data on the line segment W3-W4 is captured (step S3).
25).

Then, the same as in step S302 described later, the conversion routine is called for the width information (step S302).
326), and convert the extracted data into width information.
Next, based on this width information, a portion corresponding to a bar code table not shown, which is predetermined according to the intended use of the bar code, is converted into a code and the information is stored (step S32).
7).

Further, it is judged whether or not the code at the target location in steps S324 to S326 has been embedded (step S328). If it is not embedded in this judgment (NO), the inclination pattern counter n is incremented (step S329), and the inclination pattern number N is obtained. It is checked whether or not it is less than PAT (for example, if 10 types of inclination are prepared in the array dt in advance, set to “10”) (step S330). If less, return to step S324, and if it is more than Is S33
Move to 1.

If it is determined that the pattern has been embedded in step S328 (YES), or if the inclination pattern counter n indicates the number of inclination patterns N. If PAT or higher, the variable co
Unter is incremented, and the inclination pattern counter n is reset to "0" (step S331). It is checked whether or not the reset variable counter value exceeds the number of codes considered to be in the screen among the variable i, that is, the code words that cannot be embedded in the optimum scan (step S332). If it does not exceed (YES), the process returns to step S324, and if it exceeds (YES), it is determined whether the information described in the label of the stored code information can be completely decoded (step S333). .

If the result of this determination is that decoding is possible (YE
S), and returns to the upper routine of FIG. However, if decryption is not possible (N
O), and returns to the upper routine of FIG. 4 with the information that cannot be decoded.

Therefore, from step S324 to step S
In 332, of the codewords that could not be embedded in the optimum scan, the image data is taken in at a plurality of types of inclinations for each code that seems to be in the screen, and the code is being decided. As a result, only the necessary part is fetched, and the codeword matrix is determined with the minimum number of tries.

Next, the conversion routine to the width information called in step S302 or step S326 will be described with reference to a series of flowcharts in FIGS.

This routine is basically the same as that shown in FIGS.
Although it is the same as the routine described in step 8, only steps S95 and S355 are different. That is,
In the routine of FIGS. 17 and 18, the start code 22
The width information is WID
TH It would have been better to obtain NUM pieces, but FIG. 44 and FIG.
In this routine, all the information on the label is needed, so the width must be calculated. Therefore, in step S355, the judgment regarding j found in step S95 is removed.

It is needless to say that the present invention is not limited to the above-mentioned embodiment and various modifications can be made. For example, the two-dimensional image pickup device 5 is not limited to one using an area sensor represented by a two-dimensional CCD or an image pickup tube, but may be a combination of a one-dimensional image pickup device and a one-dimensional scan mechanism, or a photoelectric detector. A combination of two-dimensional scanning mechanisms may be used.

Further, in the above description, the label of the PDF-417 format is used as the bar code label, but the bar code label is not limited to this, and other stacked bar codes such as Code 49, JAN, etc. Dimensional barcodes are also acceptable.

Although described based on the above embodiments, the present invention includes the following inventions. (1) Image capturing means for capturing a two-dimensional image of a bar code including a bar and a space, storage means for temporarily storing the two-dimensional image obtained by the image capturing means, and the image capturing means to the storage means. Transfer means for transferring the captured image,
A recognition means for automatically recognizing the presence or absence of a barcode symbol from the two-dimensional image of the barcode obtained by the image pickup means, and a position of the barcode symbol from the two-dimensional image of the barcode obtained by the image pickup means. Position recognizing means, reading means for sequentially reading the barcode information from the two-dimensional image of the barcode obtained by the image pickup means, and decoding for decoding the barcode information from the reading means to the original information. A means for reading symbol information, comprising:

Therefore, a bar code consisting of a bar and a space is picked up as a two-dimensional image, the obtained two-dimensional image is temporarily stored, the picked-up image is transferred to the storage means, and the presence or absence of the bar code symbol is automatically detected. Recognizing, recognizing the position of the bar code symbol, sequentially reading the bar code information, and decoding the bar code information to the original information.

(2) In the first item, the recognition means recognizes the two-dimensional image of the moving barcode at a position where the barcode information can be read on the image pickup screen obtained by the image pickup means. .

Therefore, the two-dimensional image of the moving bar code is recognized at the position where the bar code information can be read on the image pickup screen obtained by the image pickup means. Therefore, the effect is that when reading a moving bar code label, it will proceed to the reading process only when the bar code label is completely inside the image, so the bar code protruding from the image will be read. There is no need to go, high-speed reading can be performed, and reliability can be improved.

(3) In the first item, the transfer means compresses and transfers the two-dimensional image picked up by the image pickup means. Therefore, the two-dimensional image picked up by the image pickup means is compressed and transferred.

Therefore, the effect is that since the target image is compressed and transferred to the frame memory 6, the number of pixels to be processed is reduced and the barcode symbol information can be read at high speed.

(4) In the above-described item 3, the transfer means transfers only one field component of the captured two-dimensional image to the storage means. Therefore, only one field component of the captured two-dimensional image is transferred to the storage means.

Therefore, the effect is that the barcode label image occupies a large proportion of the target image, and when the height of one line of the barcode label is sufficient, only the field image is transferred to the frame memory and the reading process is performed. Therefore, the bar code information can be read by an inexpensive interlaced image sensor. The field image may have either an even component or an odd component.

(5) In the above-described item (2), the transfer means transfers the two field components of the captured two-dimensional image to the storage means separately. Therefore, the two field components of the captured two-dimensional image are separately transferred to the storage means.

Therefore, the effect is that when the ratio of the barcode label image to the target image is small and the height of one line of the barcode label is not sufficient, two field images (even component and odd component) are displayed. It is possible to increase the reliability of reading by loading the image data into one frame memory separately, performing the reading process for one field image, and if that fails, performing the reading process for the other field image. is there.

(6) In the second item, the reading means performs a reading process on one field component stored by the storage means, and when the processing fails, another reading is performed on another field component stored by the storage means. A symbol information reading device for reading barcode information, including a processing operation for going to the reading process again.

Therefore, a reading operation is performed on one field component stored by the storage means, and when the processing fails, another reading operation is performed again on another field component stored by the storage means. Read barcode information.

Therefore, the same effect as the item (5) can be obtained. (7) In the second item, the transfer means transfers one scan line of the field component of the two-dimensional image picked up by the image pickup means, The scan line is transferred again after leaving one line of the storage means, and this is repeated until the image information for one field is transferred, and after the end, the image information of one line above or below that line is added to the vacant line. A symbol information reader including a processing operation for copying and transferring.

Therefore, after transferring one scan line of the field component of the two-dimensional image picked up by the image pickup means,
The scan line is transferred again after leaving one line of the storage means, and this is repeated until the image information for one field is transferred, and after the end, the image information of one line above or below that line is added to the vacant line. Includes processing operations for copying and transfers.

Therefore, the effect is that one field image is taken into the frame memory 6 at intervals of one line, the contents of the frame memory of the one line are copied to an empty line, and the reading process is performed. That is, the reliability of reading can be improved. Copying can be easily realized by hardware.

(8) In the symbol information reading device described in the third item, the transfer means adds the pixel information of an arbitrary number of the two-dimensional image picked up by the image pickup means into one pixel and transfers it.

Therefore, in the above item 3, the transfer means is
An arbitrary number of pieces of pixel information of the two-dimensional image picked up by the image pickup means are added to form one pixel and transferred. Therefore, the effect is
When the height of one line of the barcode label is sufficient, the arbitrary number of pixels is added to all the pixel information to make it 1 pixel and stored in the frame memory, so that the barcode symbol information can be processed at high speed. It is possible to read. This can also be easily realized by hardware, and also has the effect of suppressing noise.

(9) In the first item, the position recognizing means converts the pixel information of one row of the storage device into width information by an arbitrary number, and recognizes the position of a predetermined pattern defined by the kind of the barcode. Symbol information reading device.

Therefore, the pixel information of one row of the storage device is converted into the width information by an arbitrary number, and the position of the predetermined pattern determined by the type of the barcode is recognized. Therefore,
The effect is that when there is no information other than the barcode label in the target image and the start code of the PDF417 barcode label is located on the left side of the image,
This means that it is possible to detect the start code at a high speed because the width information that is sufficient for detecting the start code is obtained without converting all the pixel information fetched in line units from the frame memory into the width information.

(10) In the first item, the position recognizing means includes a processing operation for selectively selecting a row from the storing means, and recognizes the position of a predetermined pattern defined by the type of the bar code concerned. .

Therefore, the position of the predetermined pattern defined by the type of the bar code is recognized, including the processing operation of selectively selecting the line from the storage means. Therefore, the effect is that in the method of determining the row in which the pixels are taken in, a method similar to the NEWTON method for iteratively obtaining the solution of the equation is adopted, and the method of gradually narrowing the space between the rows is adopted. Moreover, it is possible to detect at high speed.

(11) In the first item, the reading means detects at least two predetermined patterns defined by the type of the bar code, calculates the inclination of the bar code label, and differs from the pattern already detected. A symbol information reading device, which includes a processing operation of detecting a predetermined pattern defined by the type of the barcode and redefining the inclination of the barcode label from both patterns, and reading the barcode information.

Therefore, at least two predetermined patterns defined by the type of the bar code are detected, the inclination of the bar code label is obtained, and the predetermined pattern defined by the type of the bar code different from the pattern already detected is determined. The barcode information is read, including the processing operations to detect and redefine the slope of the barcode label from both patterns.

Therefore, the effect is that after detecting the start code, each code position is estimated from the indicator information of the bar code label, the stop code is detected from the position close to the stop code, and the bar code is detected from both the start code and the stop code. Since the tilt of the label is redefined, a more accurate tilt can be obtained and the reliability of reading can be improved.

(12) In the first item, the reading means obtains the bar code obtained from only one information of the predetermined pattern determined by the kind of the bar code when one corner of the bar code label is missing. A symbol information reading device for reading barcode information, including a processing operation for redefining the inclination with reference to the inclination of the code label.

Therefore, when one corner of the barcode label is missing, the inclination is redefined with reference to the inclination of the barcode label obtained from the information of only one of the predetermined patterns determined by the type of the barcode. The barcode information is read, including the processing operation to perform.

The same effect as the item (12) can be obtained. (13) In the second item, the recognition means vertically divides the two-dimensional image of the storage means into three equal parts, and an arbitrary row of the top area and a bottom area determined by the size of the barcode label. The bar code label can be read when the predetermined pattern defined by the barcode type cannot be detected in any row of and the predetermined pattern defined by the barcode type is detected in any row of the central area. A symbol information reading device that recognizes the presence or absence of a bar code label, including a processing operation for determining that the bar code label exists.

Therefore, the two-dimensional image of the storage means is vertically divided into three equal parts, and the bar code is selected in any row of the top area and any row of the bottom area determined by the size of the bar code label. When the predetermined pattern defined by the type of the barcode cannot be detected and the predetermined pattern defined by the type of the barcode is detected in any row of the central area, the process of determining that the barcode label exists at the readable position Recognizes the presence or absence of barcode labels, including operations.

Therefore, the effect is that when the moving bar code label is read, the reading process is performed only when the bar code label is completely within the image.
There is no need to read a bar code that sticks out of the image, the bar code can be read at high speed, and the reliability can be improved.

(14) In the tenth item, the position recognizing means first selects the center row of the two-dimensional image in the storage means, and then the center of the upper half and the center of the lower half of the two-dimensional image. Symbol information reading device that recognizes the position of a predetermined pattern defined by the type of the barcode, including the processing operation of sequentially selecting the central line of the divided image.

Therefore, first, the row at the center of the two-dimensional image in the storage means is selected, then the rows at the center of the upper half and the center of the lower half of the two-dimensional image are selected, and so on. It includes a processing operation of sequentially selecting the center line of the image, and recognizes the position of a predetermined pattern determined by the type of the barcode.

The same effect as the item (10) can be obtained. (15) In the twelfth item, the reading means has a top two of the four corners of the two-dimensional barcode label image of the storage means.
The barcode information inclination is calculated from the corners, then the barcode label inclination is calculated from the lower two corners of the four corners, and the barcode information inclination is calculated from the two inclinations. Symbol information reader to read.

Therefore, in the twelfth item, the reading means obtains the inclination of the bar code label from the upper two corners of the two-dimensional bar code label image of the storage means, and then calculates the inclination of the bar code label. The bar code information is read by including the processing operation of obtaining the inclination of the barcode label from the lower two corners and obtaining the inclination of the barcode label from the two inclinations.

(16) In the twelfth item, the reading means obtains the inclination of the bar code label from the upper two corners of the four-dimensional corner of the two-dimensional bar code label image of the memory means, and then calculates the inclination of the bar code label. The inclination of the barcode label is calculated from the two lower corners, and if the difference between the two inclinations is greater than or equal to a predetermined threshold value, at least two predetermined patterns determined by the type of the barcode are detected. Including the processing operation that makes the value close to the calculated slope the slope of the barcode label,
A symbol information reader that reads barcode information.

Therefore, the inclination of the barcode label is calculated from the upper two corners of the four corners of the two-dimensional barcode label image of the storage means, and then the inclination of the barcode label is calculated from the lower two corners of the four corners. If the difference between the two slopes is greater than or equal to a predetermined threshold value, a value close to the slope obtained by detecting at least two predetermined patterns defined by the type of the barcode is the slope of the barcode. Read the barcode information including the processing operation.

The same effect as the item (10) can be obtained. (17) In the twelfth item, the reading means is the top two of the four corners of the two-dimensional barcode label image of the storage means.
The inclination of the barcode label is obtained from the corner, then the inclination of the barcode label is obtained from the lower two corners of the four corners, and if the difference between the two inclinations is less than a predetermined threshold value, then 2
A symbol information reading device for reading barcode information, which includes a processing operation for determining the average value of two inclinations as the inclination of the barcode label.

Therefore, the inclination of the bar code label is calculated from the upper two corners of the four corners of the two-dimensional bar code label image of the storage means, and then the inclination of the bar code label is calculated from the lower two corners of the four corners. If the difference between the two slopes is less than a predetermined threshold value, the bar code information is read, including the processing operation of setting the average value of the two slopes as the slope of the barcode label.

(18) In the sixth item, the reading means includes a processing operation of using the position information of the bar code label of the previous field screen when going to the reading processing again.
A symbol information reader that reads barcode information.

Therefore, the bar code information is read by including the processing operation using the position information of the bar code label on the previous field screen when going to the reading process again. Therefore, the effect is that since the position information of the bar code label on the previous field screen can be used when the reading process is performed again, it is not necessary to detect the label again, and high-speed reading of the bar code information can be realized. . The new bar code label position can be calculated from the label speed and the scanning speed of the imager. Further, when the speed of the barcode label is sufficiently negligible as compared with the scanning speed of the image pickup apparatus, the position information of the barcode label on the previous field screen can be used as it is.

[0225]

As described in detail above, according to the present invention, it is possible to provide a symbol information reading device which can reliably read barcode symbol information of a moving two-dimensional barcode at high speed.

[Brief description of drawings]

FIG. 1 is a diagram showing a schematic configuration of a symbol information reading device as a first embodiment.

FIG. 2 shows a PDF-
It is a figure which shows the label structure of 417.

FIG. 3 is a view showing a PDF-pixel arrangement in a pixel array of a frame memory.
It is a figure which shows the schematic diagram which projected the label image of 417.

FIG. 4 is a flowchart for explaining label detection, label information reading, and decoding in the data processing device.

FIG. 5 is a diagram for explaining that only a field screen is captured.

FIG. 6 is a flowchart illustrating an image capturing routine.

[Fig. 7] Two field screens (even component and odd component)
FIG. 6 is a diagram for explaining that the capture of

FIG. 8 is a diagram for explaining that four pixels are added to form one pixel.

FIG. 9 is a flowchart for explaining barcode label detection.

FIG. 10 is a diagram showing a projected image for explaining barcode label detection.

FIG. 11 is a diagram for explaining the inclination of the barcode label defined on the frame memory.

FIG. 12 is a flowchart for explaining detection within a barcode label screen.

FIG. 13 is a diagram showing a projected image of a barcode label for explaining detection within a barcode label screen.

FIG. 14 is a flowchart illustrating a scanning method.

FIG. 15 is a diagram showing a barcode label projection image for explaining a scanning method.

FIG. 16 is a flowchart for explaining start code detection.

FIG. 17 is the first half of a flowchart for explaining a conversion method into width information.

FIG. 18 is the second half of the flowchart for explaining the conversion method into the width information.

FIG. 19 is a diagram for explaining selection of peak values.

FIG. 20 is a diagram for explaining a method of calculating a distance between peaks.

FIG. 21 is a flowchart for explaining normalization.

FIG. 22 is a flowchart for explaining detection of the maximum value and minimum value of Y coordinates.

FIG. 23 is a flowchart for explaining detection of a label range.

FIG. 24 is a flowchart for explaining detection of label inclination.

FIG. 25 is a flowchart for explaining threshold determination.

26A is a diagram for explaining a threshold value calculation line, and FIG. 26B is a diagram showing a differential waveform of a start code.

FIG. 27 is a flowchart for explaining determination of indicator information.

FIG. 28 is a diagram showing a bar code label projected image on a frame memory for determining indicator information.

FIG. 29 is a flowchart for explaining redefinition of the inclination of the barcode label.

FIG. 30 is a diagram showing a barcode label projection image for explaining redefinition of the inclination of the barcode label.

FIG. 31 is a flowchart for explaining stop code detection.

FIG. 32 is a flowchart for explaining detection of a barcode label range.

FIG. 33 is a flowchart for explaining the recalculation of the inclination of the barcode label.

FIG. 34 is a diagram for explaining a defect of a barcode label.

FIG. 35 is a diagram showing a barcode label projection image for explaining calculation of a starting point and an ending point sequence of a barcode label.

FIG. 36 is a flowchart for explaining determination of a scan equation.

FIG. 37 is a flowchart for explaining an embedding check.

FIG. 38 is a diagram showing a barcode label projected image for explaining the embedding check.

FIG. 39 is a flowchart for explaining ROW detection.

FIG. 40 is a flowchart for explaining ROW number detection.

FIG. 41 is a flowchart for explaining an optimum scan.

FIG. 42 is a diagram showing a bar code label projected image on a frame memory for explaining the optimum scanning.

FIG. 43 is a flowchart illustrating an intelligent scan.

FIG. 44 is the first half of a flowchart for explaining a method of converting to width information.

FIG. 45 is the second half of the flowchart for explaining the method of converting to width information.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Article, 2 ... Bar code label, 3 ... Conveying system, 4 ... Imaging optical system, 5 ... Two-dimensional imaging part, 6 ... Frame memory, 7
... data processing device, 8 ... label detecting unit, 9 ... reading unit,
10 ... Decode determination unit, 11 ... Decoding unit, 12 ... Memory,
13 ... Control unit, 14 ... Position detection unit, 15 ... Inclination detection unit,
16 ... Host device.

Claims (3)

[Claims] 1. An image pickup means for picking up a two-dimensional image of a bar code which is composed of a bar and a space and which represents arbitrary information as a pattern, and a storage means for temporarily storing the two-dimensional images sequentially obtained by the image pickup means. A transfer means for transferring a two-dimensional image from the imaging means to the storage means; and a recognition means for sequentially reading out the two-dimensional images stored in the storage means and recognizing the presence or absence of a bar code symbol within the imaging range. A position detecting unit that detects the position and inclination of the barcode recognized by the recognizing unit in the imaging range and detects the scanning direction of the barcode; and a bar based on the scanning direction detected by the position detecting unit. A reading unit that sequentially reads information of the barcode from a two-dimensional image of the code, and decodes the barcode information from the reading unit into the arbitrary information. A symbol information reading apparatus comprising: a decoding unit. 2. The symbol information reading device according to claim 1, wherein the transfer unit compresses and transfers the two-dimensional image captured by the image capturing unit. 3. The transfer means transfers one scan line of a field component of a two-dimensional image picked up by the image pickup means, vacates one line of the storage means, and transfers the scan line again, and one field worth The process is repeated until the image information is transferred, and after the process is completed, a processing operation of copying the image information of one line above or below the line to a vacant line is included and transferred.
The described symbol information reader.