JP7039918B2 - Segment management method - Google Patents

Description

The present invention relates to a method of managing segments used in a shield tunneling method.

Patent Document 1 discloses a segment management method used in the shield tunneling method. Specifically, the barcode is pasted on the segment, the barcode is read by the barcode reader, and the identification information represented by the barcode is input to the personal computer. The identification information is information for identifying individual segments.

Japanese Unexamined Patent Publication No. 2002-2146

By the way, segments are usually stored outdoors. Therefore, the barcode is peeled off from the segment due to wind and rain, and the information is lost.
Further, as long as the barcode is a code, the information represented by the barcode cannot be understood by the operator. Therefore, characters are printed on the segment separately from the barcode so that the operator can understand the information represented by the barcode. However, the information is doubly attached to the segment.

Therefore, the present invention has been made in view of the above circumstances, and the problem to be solved by the present invention is to make it difficult for the information assigned to the segment to disappear due to wind and rain, and to provide such information. It is to avoid having to attach to the segment heavily.

In order to solve the above problems, in the segment management method for managing the segment by having a computer read the allocation information of the segment represented by a plurality of characters printed on the segment used in the shield method. The allocation information is divided into at least items of type information, assembly method information, taper information, and serial number information, and the plurality of characters printed in the segment are grouped into a plurality of areas for each item . The plurality of characters are converted into a plurality of character codes by inputting an image containing characters into the computer and causing the computer to perform character recognition processing of the image, and at the same time, the computer is subjected to the plurality of characters in the character recognition processing. The character code of is divided into the areas, a character code string or a character code is generated in the computer for each item, and the plurality of character codes are registered in the computer as a character code string or a character code for each item. do.

According to the above, by printing a plurality of characters on the segment, the allocation information represented by these characters is attached to the segment. The printed characters are hard to leave the segment and therefore the allocation information is hard to lose. Moreover, since what is printed on the segment is a character, the allocation information represented by these characters can be understood by the operator.
An image containing a plurality of characters is input to a computer, and the computer is subjected to OCR processing of the image to convert the plurality of characters into a plurality of character codes. As a result, the allocation information represented by these characters is registered in the computer. Therefore, it is not necessary to attach the barcode to the segment.
Type information, assembling method information, taper information, and serial number information as items of allocation information can be acquired as a character code string or a character code.

Preferably, the plurality of characters printed on the segment are grouped into a plurality of lines, and in the character recognition process, the computer is made to cut out the image into a line-by-line area, and the computer is made to perform the line. Each area is cut out into an area for each character, and the character code is determined by collating a plurality of collation character images associated with the character code with the area for each character. The plurality of collation character images are images in which the characters used in each line are shown.

According to the above, since a plurality of collation character images are images of characters used in each line, the recognition rate of the characters in each line is high.

Preferably, the plurality of characters printed on the segment are grouped into a plurality of areas according to the item of the allocation information, and the plurality of character codes are divided into the plurality of areas by the computer during the character recognition process. At the same time, the computer is made to generate a character code string or a character code for each of the items.

According to the present invention, the allocation information assigned to the segment is unlikely to be lost due to wind and rain, and it is not necessary to duplicate such information to the segment.

FIG. 1 is a segment management process diagram. FIG. 2 is an explanatory diagram showing a segment and a plurality of characters printed on the segment. FIG. 3 is a block diagram of the segment management system.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, although the embodiments described below are provided with various technically preferable limitations for carrying out the present invention, the scope of the present invention is not limited to the following embodiments and illustrated examples.

1. 1. Segment management process FIG. 1 is a segment 1 management process diagram used in the shield tunneling method.
As shown in FIG. 1, the segment 1 is manufactured in the factory 81, and after the manufacturing inspection is performed, the manufactured segment 1 is stored in the stockyard 82 of the factory 81. Manage segment 1 for storage. After that, the segment 1 is taken out from the stockyard 82, the segment 1 is inspected, and then the shipment is prepared. The segment 1 ready for shipment is loaded on the freight vehicle 83 and transported to the construction site 84. At the construction site 84, after receiving and inspecting the segment 1, the segment 1 is unloaded from the freight vehicle 83 and temporarily stored in the stockyard 85 of the construction site 84. The segments 1 are carried from the stockyard 85 to the temporary storage place 87 at the bottom of the shaft 86, the type of the segment 1 is confirmed, and then the segment 1 is loaded on the trolley 89. In the shield tunnel 88 whose construction has been completed, the segment 1 is transported from the temporary storage place 87 to the assembly site near the tail of the shield machine 90 by the trolley 89. Then, the segment 1 is unloaded from the trolley 89. At the end of the shield tunnel 88, the shield machine 90 is sequentially advanced while excavating the face by the shield machine 90. At that time, the segment 1 is laid on the inner peripheral surface of the excavated tunnel by assembling the segment 1 in a ring shape on the tail of the shield machine 90 using an elector. A unique number (for example, a serial number or a serial number) is assigned to each segment 1.

2. 2. Segment allocation information For inspection and management of segment 1, the allocation information assigned to each segment 1 is used. The allocation information may be unique to each segment 1 or may be unique to each type of segment 1. The allocation information unique to each segment 1 is called unique identification information, and the allocation information unique to each type of segment 1 is called type identification information. The content (value) of the allocation information assigned to each segment 1 is represented by the characters printed on the surface of the segment 1. A character is a character code assigned on a computer. Therefore, punctuation marks and symbols to which character codes are assigned are also types of characters. Punctuation marks are a general term for various symbol types used in language descriptions, such as punctuation marks, question marks, parentheses, accents, and emphasis marks.

FIG. 2 shows a display example of the allocation information printed on the surface of the segment 1. As shown in FIG. 2, a character string displaying the contents of the allocation information is attached to the predetermined print area 2 on the surface of the segment 1 by printing. These characters are arranged horizontally in two lines. That is, a plurality of characters are arranged in the horizontal direction in the print area 3 on the first line, and a plurality of characters are arranged in the horizontal direction in the print area 7 on the second line.

The print area 3 on the first line is divided into a central area 5, an area 4 on the line head side of the central area 5, and an area 6 on the line end side of the central area 5, depending on the character type. That is, the character type used in the central area 5 is different from the character type used in the area 4 at the beginning of the line and the character type used in the area 6 on the end side of the line. Specifically, the character type in the central area 5 is Chinese characters, the character type in the area 4 at the beginning of the line is an alphabetic character, an Arabic numeral or a punctuation mark, and the character type in the area 6 at the end of the line is an alphabetic character, an arithmetic symbol or a punctuation mark. Is. In particular, the characters used in the central area 5 are limited to "A" and "B", and the characters used in the area 4 on the beginning of the line are "A", "B", "K", and " It is limited to "1" to "9", "-" hyphen "", and the characters used in the area 6 on the end of the line are "L", "R", "S", "(" opening parentheses "". , ")" Closed parentheses "", "=" Equal sign "". In this specification, what is written in parentheses in double brackets means the content (value) represented by characters, data, information, or the like.

The character type in the print area 7 on the second line is an Arabic numeral. That is, the characters used in the print area 7 are limited to "0" to "9".

The plurality of characters printed in each segment 1 are grouped into areas 4 to 6 for each item. The items of the allocation information assigned to each segment 1 include "type", "assembly method", "taper", and "serial number". That is, the allocation information is divided into type information, assembly method information, taper information, and serial number information.
The content of the type information is represented by a character or a character string printed in the area 4 on the line head side of the first line. The type information includes at least information in which the type of segment 1 classified by shape is expressed by characters or character strings, and may also include information in which the order in the circumferential direction in which the segment 1 is incorporated is expressed by characters or character strings. The types of segment 1 include, for example, A type, B type, and K type.

The content of the knitting method information is represented by a character or a character string printed in the central area 5 of the first line. The composition method information is information that expresses the composition method of segment 1 by characters or character strings. There are two ways to assemble segment 1, instep and second.

The content of the taper information is represented by a character or a character string printed in the area 6 on the end side of the first line. The taper information is information expressing the taper of the shield ring assembled from the segment 1 by a character or a character string. The taper includes a right taper, a left taper, and a straight.

The content of the serial number information is represented by a character string printed in the area 7 on the second line. The serial number information is information in which the serial number of segment 1 is expressed by a character or a character string.

As described above, by printing a plurality of characters on the segment 1, the allocation information represented by these characters is attached to the segment 1. The characters printed in segment 1 are hard to be separated due to wind and rain, and the allocation information is hard to be lost.
Further, since what is printed in segment 1 is a character, the information / meaning represented by these characters, that is, the content of the allocation information can be understood by the worker.

3. 3. Management system As shown in FIG. 1, when inspecting and managing the segment 1, the image data showing the contents of the allocation information printed in the segment 1 is taken into the terminal computer 21 by the image input device 25. The character recognition function of the terminal computer 21, for example, the OCR (Optical Character Recognition) function converts the image data into a character code, and causes the terminal computer 21 to recognize the content of the allocation information as the character code. In this way, the allocation information is registered in the terminal computer 21 to inspect and manage the segment 1. For example, the registration time of the allocation information (that is, the product inspection time, the shipping time, the construction site acceptance time, the shaft loading time, the temporary storage site carry-in time, the assembly site carry-in time) and the position information are associated with the allocation information in the terminal computer 21. By registering, determining whether or not the registered allocation information matches the predetermined regular information, and transmitting the allocation information from the terminal computer 21 to the host computer 51 and registering it in the host computer 51. , Performs segment 1 inspection and management. By such image input processing and OCR processing, the allocation information represented by the characters printed in the segment 1 is registered in the terminal computer 21, so that it is not necessary to separately attach the barcode to the segment 1.

A management system including a host computer 51, a terminal computer 21, and the like will be described in detail with reference to FIG.
The host computer 51 includes a CPU, GPU, RAM, ROM, system bus, hardware interface, and the like. A storage unit 52, a display device 53, an input device 54, and a communication module 55 are connected to the host computer 51. The storage unit 52 is a storage device including a semiconductor memory, a hard disk drive, or the like. The display device 23 is a display device that displays according to a video signal output by the host computer 51. The input device 54 is a keyboard and a pointing device. The communication module 55 communicates with the communication module 26 of the terminal computer 21 according to a predetermined communication standard. Communication is realized between the terminal computer 21 and the host computer 51 by the communication modules 55 and 26.

The terminal computer 21 has a CPU, a GPU, a RAM, a ROM, a system bus, a hardware interface, and the like. A storage unit 22, a display device 23, an input device 24, an image input device 25, and a communication module 26 are connected to the terminal computer 21.

The storage unit 22 is a storage device including a semiconductor memory, a hard disk drive, or the like. The storage unit 22 may be built in the terminal computer 21 or externally attached to the terminal computer 21. The storage unit 22 stores a first collation character image group 31 and a second collation character image group 32 that can be read by the terminal computer 21.

The first collation character image group 31 is composed of a plurality of raster format character images in which characters and character backgrounds are represented by different pixel values (colors). One character is copied for each character image. The same applies to the second collation character image group 32. The character images of the collation character image groups 31 and 32 may be feature-extracted.
The characters copied to each character image of the first collation character image group 31 are the characters used in the print area 3 of the first line. Each character image of the first collation character image group 31 is associated with a character code obtained by encoding characters. The characters encoded in the character code are the characters used in the print area 3 on the first line.

The characters copied to each character image of the second collation character image group 32 are the characters used in the print area 7 on the second line. Each character image of the second collation character image group 32 is associated with a character code obtained by encoding characters. The characters encoded in the character code are the characters used in the print area 7 on the second line.

The collation character image groups 31 and 32 may be stored in the storage unit 22 in advance, or may be downloaded from the host computer 51 to the terminal computer 21 and stored in the storage unit 22.

The storage unit 22 stores a program 30 that can be executed by the terminal computer 21. The program 30 causes the terminal computer 21 to execute the processing described later.

The display device 23 is, for example, a liquid crystal display device, an organic EL display device, or a projector. The terminal computer 21 generates a video signal by arithmetic processing and outputs the video signal to the display device 23. Then, the screen according to the video signal is displayed on the display device 23.

The input device 24 is, for example, a switch, a keyboard or a pointing device, or a combination thereof. When the input device 24 is operated, the input device 24 outputs a signal corresponding to the operation content to the terminal computer 21.

The image input device 25 is, for example, a digital camera (imaging device) or an image scanner. The image input device 25 converts an optical image including the print area 2 of the segment 1 into a digital image (hereinafter referred to as an input image) by photoelectric conversion, and outputs the optical image to the terminal computer 21. The terminal computer 21 acquires an input image from the image input device 25 and stores it in the storage unit 22. The operator uses the image input device 25 to input an image twice or more per segment 1. Therefore, a plurality of input images for one segment 1 are transferred from the image input device 25 to the terminal computer 21. The input image may be recorded on a portable storage medium that can be attached to and detached from the image input device 25. In that case, when the reader / writer is connected to the terminal computer 21 and the portable storage medium is attached to the reader / writer, the terminal computer 21 can read the input image recorded on the portable storage medium.

Subsequently, the process of causing the terminal computer 21 to execute the program 30 will be described.
First, the terminal computer 21 collates the input image recorded in the storage unit 22 or the portable storage medium with the first collation character image group 31, and the character string printed in the print area 3 on the first line. Recognize. Further, the terminal computer 21 recognizes the character string printed in the print area 7 on the second line by collating the input image with the second collation character image group 32. Hereinafter, a specific description will be given.

First, the terminal computer 21 recognizes two areas (a region corresponding to the print area 3 and an area corresponding to the print area 7) in the input image by performing image processing and layout analysis of the input image. Then, the terminal computer 21 cuts out these two areas from the input image. Of the two areas, the area corresponding to the print area 3 is referred to as the area of the first line, and the area corresponding to the print area 7 is referred to as the area of the second line.

Next, the terminal computer 21 cuts out the cut out area of the first line into a character-by-character area.
Next, the terminal computer 21 performs a matching process for each character area. That is, the terminal computer 21 collates the area for each character with each image of the first collation character image group 31 after performing the feature extraction process of the area for each character. Then, the terminal computer 21 determines the character image most suitable for the area for each character and the character code corresponding to them, and generates a string of those character codes. Hereinafter, this character code string is referred to as a character code string on the first line. As described above, the terminal computer 21 recognizes the character string printed in the print area 3 on the first line.
A neural network constructed by deep learning may be used for the character recognition process of the area for each character. Here, when a large number of input data (character images) and a large number of output data (character codes) are read into the terminal computer 21, the terminal computer 21 extracts the image feature amount and classifies the image feature amount based on the image feature amount. As a result, the terminal computer 21 constructs a convolutional type or recursive type neural network, and causes the terminal computer 21 to perform deep learning type machine learning. In the character recognition process, the terminal computer 21 inputs an area for each character into the neural network and acquires a character code as an output of the neural network. At this time, the terminal computer 21 deepens the intermediate layer of the neural network by using the character recognition result converted from the area for each character into the character code. As a result, character recognition processing and deep learning are performed in parallel.

Similarly for the cut out second line area (area corresponding to the print area 7), the terminal computer 21 collates the area for each character on the second line with each character image of the second collation character image group 32. do. As a result, the terminal computer 21 recognizes the character string printed in the print area 7 on the second line and generates the character code string of the character string. Hereinafter, this character code string is referred to as a character code string on the second line.
Further, the terminal computer 21 recognizes the item corresponding to the character code string on the second line as the "serial number".

After recognizing the character strings in the print areas 3 and 7 as described above, the terminal computer 21 has a character code corresponding to the character in the central area 5 from the character code string on the first line (hereinafter, the character code in the center). ) Is specified. Here, as described above, the character type used in the central region 5 is different from the character type used in the other regions 4 and 6. In particular, the characters used in the central region 5 are limited to specific characters (that is, "A" or "B"). Therefore, the central character code is specified by the terminal computer 21.

Then, the terminal computer 21 decomposes the character code string on the first line into three sets. That is, the terminal computer 21 uses the character code string of the first line as the central character code, the character code or character code string at the beginning of the line, and the character code or character code string at the end of the line. Disassemble. Since the central character code is specified by the terminal computer 21, such decomposition can be performed.
Further, the terminal computer 21 recognizes an item corresponding to the recognized character code or character code string at the beginning of the line as a "type". Further, the terminal computer 21 recognizes the item corresponding to the recognized central character code as "how to assemble". Further, the terminal computer 21 recognizes the recognized character code at the end of the line or the item corresponding to the character code string as "taper".

Next, the terminal computer 21 represents the recognized character code or character code string at the beginning of the line, the unique identifier unique to each segment 1, the item identifier unique to each item, and the time when the character of the input image is recognized. The time stamp is associated with the time stamp and recorded in the storage unit 22. Here, a unique number (for example, a serial number or a serial number) is assigned to each segment 1, and the value of the unique identifier is assumed to be equal to that number. Further, the item here is a "type", and the item identifier is a code representing the "type". For example, if the value of the code is "1", the item is "type".

Further, the terminal computer 21 records the character code in the center of the character code string on the first line, the unique identifier, the item identifier, and the time stamp in association with each other in the storage unit 22. The item here is "how to assemble", and the item identifier represents "how to assemble". The item here is "how to assemble", and the item identifier is a code representing "how to assemble". For example, if the value of the code is "2", the item is "how to assemble".

Further, the terminal computer 21 records the character code or the character code string at the end of the line, the unique identifier, the item identifier, and the time stamp in the storage unit 22 in association with each other in the character code string of the first line. The item here is "taper", and the item identifier is a code representing "taper". For example, if the value of the code is "3", the item is "taper".

Further, the terminal computer 21 records the character code string on the second line, the unique identifier, the item identifier, and the time stamp in association with each other in the storage unit 22. The item here is a "serial number", and the item identifier is a code representing the "serial number". For example, if the value of the code is "4", the item is "serial number".

As described above, the information (type information, assembly method information, taper information and serial number information) regarding the segment 1 in which the image of the print area 2 is input by the image input device 25 is registered in the storage unit 22 of the terminal computer 21 and managed. Will be done.

The terminal computer 21 repeats the same process as acquiring the information (type information, assembly method information, taper information, and serial number information) related to the segment 1 by the OCR process described above a plurality of times. Specifically, the terminal computer 21 performs OCR processing on another input image of the same segment 1 to acquire information about the segment 1. Therefore, for the same segment 1, information about the segment 1 is recorded in the plurality of storage units 22. Then, the terminal computer 21 collates a plurality of information about the segment 1 with each other and determines whether or not all the information is the same. If the plurality of information regarding the segment 1 are all the same, the terminal computer 21 determines the information as authentic and causes the display device 23 to display the authenticity. On the other hand, if there is any difference in the plurality of information regarding the segment 1, the terminal computer 21 determines the information as an error and causes the display device 23 to display the error. After that, the operator again inputs the image twice or more per segment 1 using the image input device 25.

Further, the terminal computer 21 transmits each of the associated information as described above to the host computer 51. Then, the host computer 51 receives the information from the terminal computer 21 and records it in the storage unit 52. As a result, the information about the segment 1 is registered and managed in the storage unit 52 of the host computer 51.

4. Advantageous effect (1) Since the characters printed on the segment 1 are hard to be separated due to wind and rain, the allocation information assigned to the segment 1 is hard to be lost.

(2) Since the plurality of characters printed in the segment 1 represent the contents of the allocation information, the worker can recognize the contents of the allocation information.

(3) In each process from the manufacture of the segment 1 to the assembly, the input image is input to the terminal computer 21 by the image input device 25, and the content of the allocation information is registered in the terminal computer 21 by the OCR function of the terminal computer 21. Therefore, traceability of the distribution channel of segment 1, that is, traceability can be realized.

(4) Since the characters copied to each image of the first collation character image group 31 are limited to the characters that can be used in the print area 3 of the first line, the characters in the print area 3 of the first line Character recognition rate is high. Similarly, the character recognition rate in the print area 7 on the second line is high. In particular, since the character type in the print area 7 on the second line is an Arabic numeral, the character recognition rate is further improved.

5. Modifications Although the embodiments for carrying out the present invention have been described above, the above-described embodiments are for facilitating the understanding of the present invention, and are not intended to limit the interpretation of the present invention. Further, the present invention can be modified or improved without departing from the spirit thereof, and the present invention also includes an equivalent thereof. The changes from the above embodiments will be described below. The changes described below may be applied in combination as much as possible.

(1) The allocation information of segment 1 may be divided into at least type information, assembly method information, taper information, and serial number information. That is, the allocation information includes type information, assembly method information, taper information, and serial number information, as well as other information (for example, type information, usage information, manufacturing date information, manufacturing factory information, manufacturer information, manufacturing time). It may be divided into items (quality information).

(2) In the above embodiment, the terminal computer 21 has an OCR function. The host computer 51 may also have an OCR function. In this case, when the input image is input to the terminal computer 21 by the image input device 25, the terminal computer 21 associates the terminal identifier unique to the terminal computer 21 with the input image, the image input time, and the position information, and these terminals. The identifier, the input image, the image input time, and the position information are transmitted to the host computer 51. Then, the host computer 51 uses the OCR function to obtain the character code in the center of the first line from the input image, the character code or character code string at the beginning of the first line, and the character code or character code string at the end of the first line. And the character code string of the second line are generated. Then, the host computer 51 records these character codes or character code strings in the storage unit 52 in association with the unique identifier, the item identifier, the terminal identifier, the time stamp (image input time), and the position information.

1 ... Segment 2 ... Print area 3 ... 1st line print area 4 ... 1st line head side area 5 ... 1st line center area 6 ... 1st line end side area 7 ... 2nd line printing Area 21 ... Terminal computer 22 ... Storage unit 25 ... Image input device 30 ... Program 31 ... First collation character image group 32 ... Second collation character image group 51 ... Host computer 52 ... Storage unit 88 ... Shield tunnel 90 … Shield machine

Claims (2)

In the segment management method of managing the segment by having a computer read the allocation information of the segment represented by a plurality of characters printed on the segment used in the shield method.
The allocation information is divided into at least items of type information, knitting method information, taper information, and serial number information, and the plurality of characters printed in the segment are grouped into a plurality of areas for each item.
An image in which the plurality of characters are captured is input to the computer, and the computer performs character recognition processing on the image to convert the plurality of characters into a plurality of character codes, and the computer performs the character recognition processing. Is divided into the plurality of character codes for each area, a character code string or a character code is generated by the computer for each item, and the plurality of character codes are used as a character code string or a character code for each item. Segment management method to register in the computer. The plurality of characters printed on the segment are grouped into a plurality of lines.
In the character recognition process, the computer is made to cut out the image into a line-by-line area, and the computer is made to cut out the line-by-line area into a character-by-character area. Let the computer perform line by line to determine the character code by collating the image with the area for each character.
The segment management method according to claim 1, wherein the plurality of collation character images are images in which characters used in each line are captured.

Images