Terminal, system, method and program for recognizing container number

Abstract

To provide a terminal, a system, a method and a program for reading and recognizing a container number when the container number is vertically written.SOLUTION: A terminal 1 for recognizing a container number comprises: an acquisition unit which acquires an image including the whole of the vertically written container number displayed on a container; a detection unit which detects a region of the whole of the container number in the acquired image and detects each character from the top character to the bottom character in the region; an inclination correction unit which determines an inclination from coordinates of the top character and the bottom character to correct the inclination; a size correction unit which calculates an average value of size of each of the detected characters to correct the size of each character to the average value; and a recognition unit which connects each of the characters with corrected size in a horizontal direction to recognize each character as the container number.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a terminal, a system, a method, and a program for recognizing a container number.

Background Art

[0002] Internationally circulated marine containers are uniquely assigned an administrative code called a container number, which is displayed on the side of the container. The container number indicates the owner of the container, the serial number, the nationality code, the structural classification, etc., and is standardized internationally. Therefore, in intermodal transportation that combines different transportation agencies such as railways, trucks, ships, and airplanes, it is possible to transport products without ever taking them out of the container or reloading them onto a loading platform or pallet.

[0003] In such intermodal transportation, since there are many and complex related parties, it is important to manage the location of the container. Therefore, at locations where containers are placed, such as factories, bonded warehouses, ports, container yards, transportation agencies, and trading company warehouses, it is possible to manage the position and status of the container by reading the container number displayed vertically or horizontally on the side of the container.

[0004] For this reason, in recent years, for containers with the container number displayed horizontally, a technology has been used to acquire an image of the container number, recognize the acquired image, and identify the container number (Patent Document 1).

[0005] Also, conventionally, there is a technology for recognizing and reading a character string from an image in which a vertically written character string is shown (Non-Patent Document 1).

Prior Art Documents

Patent Documents

[0006]

Patent Document 1

Non-Patent Document

[0007]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0008] By the way, in the technology of Patent Document 1, since container numbers displayed in vertical writing cannot be recognized by character recognition, an operator has to input them manually.

[0009] Also, in the technology of Non-Patent Document 1, when a vertically written character string is straight in an oblique direction, the vertically written character string cannot be recognized.

[0010] Therefore, the inventors of the present invention focused on the fact that when performing character recognition, by repeating the process of detecting each character for the number of characters in the vertical direction, all vertically written characters can be detected, and even if the characters are not aligned straight in a certain direction, they can be recognized.

[0011] In view of these problems, an object of the present invention is to provide a terminal, a system, a method, and a program for recognizing a container number that can be read when the container number is written vertically.

Means for Solving the Problems

[0012] The present invention provides the following means for solving the problems.

[0013] According to a first feature of the present invention, a terminal for recognizing a container number, An acquisition unit that acquires an image including the entire vertically written container number displayed on the container; In the acquired image, a detection unit that detects the area of the entire container number and detects each character from the topmost character to the bottommost character in the area; An inclination correction unit that determines the inclination from the coordinates of the topmost character and the bottommost character and corrects the inclination; A size correction unit that calculates the average value of the sizes of the detected characters and corrects the sizes of the characters to the average value; A recognition unit that combines the characters with the corrected sizes side by side and recognizes the characters as the container number; Provided is a terminal including the above.

[0014] According to the first feature of the present invention, even if the container number is displayed vertically on the container, the container number can be read, so the operator does not need to handwrite the container number on paper or input it into the terminal. Therefore, similar to the case where the container number is displayed horizontally on the container, the location management of the container becomes easy.

[0015] The second feature of the present invention is the invention according to the first feature, Provided is a terminal further including a display unit that displays the recognized container number horizontally.

[0016] According to the second feature of the present invention, even if the container number is displayed vertically on the container, on the terminal, the container number is displayed horizontally, so it becomes possible to uniformly manage all the container numbers of the containers to be managed in horizontal writing.

[0017] The third feature of the present invention is the invention according to the first feature, Provided is a terminal further including a learning unit that learns the area of the entire container number and each character in the area to create a learned model.

[0018] According to the third feature of the present invention, by learning the feature amount specific to the entire region of the container number and the feature amount specific to each character from the topmost character to the bottommost character in the region, and creating a learned model, it becomes possible to improve the automatic detection accuracy of the entire region of the container number and each character in the region.

[0019] The fourth feature of the present invention is an invention related to the third feature, Based on the learned model, in the acquired image, a terminal is provided that further includes an inference unit that infers the entire region of the container number and infers each character from the topmost character to the bottommost character in the region.

[0020] According to the fifth feature of the present invention, even when an image including the entire container number is captured at a distance by a fixed camera or the like, character recognition can be performed from the captured image using learning data, so that more accurate character recognition becomes possible.

[0021] Although the present invention is in the category of terminals, it also exhibits similar actions and effects according to the category in other categories such as systems.

Effects of the Invention

[0022] According to the present invention, it becomes possible to provide a terminal, a system, a method, and a program for recognizing a container number that can be read when the container number is written vertically.

Brief Description of the Drawings

[0023]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Mode for Carrying Out the Invention

[0024] Hereinafter, the best mode for carrying out the present invention will be described with reference to the drawings. Note that this is merely an example, and the technical scope of the present invention is not limited thereto.

[0025] [Outline of Terminal 1 for Recognizing Container Number] The outline of the terminal 1 for recognizing the container number, which is an embodiment of the present invention, will be described with reference to FIG. 1. FIG. 1 is a diagram for explaining the outline of the terminal 1 for recognizing the container number, which is an embodiment of the present invention.

[0026] The terminal 1 for recognizing the container number is, for example, a mobile terminal such as a handy terminal, a smartphone, a tablet terminal, etc., a wearable terminal such as a head-mounted display like smart glasses or a smartwatch, etc., and is equipped with an imaging device for taking images such as color videos and / or still images of a camera, etc.

[0027] The terminal 1 may be realized by, for example, one terminal device, may be realized by a plurality of terminal devices, or may be realized by a virtual device such as a cloud computer.

[0028] Next, an outline of the processing executed by the terminal 1 for recognizing the container number will be described. First, the terminal 1 acquires an image including the entire vertically-written container number displayed on the container (step S1). Specifically, the terminal 1 acquires an image in which the container number displayed on the container photographed by the user using the terminal 1 is shown.

[0029] Next, in the acquired image, the terminal 1 detects the area 100 of the entire container number, and detects each character 110 from the topmost character to the bottommost character in the area (step S2). Specifically, the terminal 1 detects the area including the entire vertically-written container number shown in the acquired image, and detects and extracts the characters included in the detected area one by one from the topmost to the bottommost. The method for detecting and extracting characters is not particularly limited.

[0030] Also, in step S2, the terminal 1 may infer the area 100 of the entire container number and infer each character 110 from the topmost character to the bottommost character in the area 100 in the acquired image based on the learned model 200. Specifically, the terminal 1 infers and extracts the area 100 of the entire container number shown in the acquired image based on the learned model 200 created by machine learning in advance, and infers each character 110 included in the extracted image. The method for inferring and extracting the area 100 of the entire container number and each character 110 is not particularly limited. Machine learning will be described later.

[0031] In step S2, the terminal 1 may relearn, by machine learning, the feature amount specific to the entire area 100 of the detected or inferred container numbers and the feature amount specific to each detected or inferred character 110 based on the learned model 200 and the detection result or inference result, and update the learned model 200. Note that this relearning by machine learning may be performed at any timing after step S2, or may be performed when the terminal 1 receives an input from the user. Machine learning will be described later.

[0032] Next, the terminal 1 determines the inclination from the coordinates of the uppermost and lowermost characters and corrects the inclination (step S3). Specifically, the terminal 1 obtains the inclination of the character string of the container number from the uppermost and lowermost characters, and corrects the character string so that the character string is vertically arranged vertically in two-dimensional coordinates with the center of the character string as the origin.

[0033] Next, the terminal 1 calculates the average value of the sizes of the detected characters and corrects the sizes of the characters to the average value (step S4). Specifically, the terminal 1 calculates the average value of the heights of the recognized characters, and enlarges or reduces each character so that the height of each character becomes the average value.

[0034] Next, the terminal 1 combines the characters with corrected sizes side by side and recognizes the container number (step S5). Specifically, the terminal 1 transposes the vertically arranged character string with corrected size to make the character string horizontal, and recognizes each character as character data. The terminal 1 may output the recognized character data horizontally on its display unit 17. Note that the method for recognizing character data is not particularly limited.

[0035] The above is an outline of the processing executed by the terminal 1 for recognizing the container number.

[0036] [System configuration of terminal 1 for recognizing container number] Based on FIG. 2, the system configuration of the terminal 1 for recognizing the container numbers of the present embodiment will be described.

[0037] The terminal 1 may be implemented by, for example, one terminal device or a plurality of terminal devices.

[0038] The terminal 1 is, for example, a mobile terminal such as a handy terminal, a smartphone, a tablet terminal, etc., or a wearable terminal such as a head-mounted display like smart glasses or a smartwatch, and includes an imaging device for capturing images such as color videos and / or still images with a camera or the like.

[0039] As a control unit and a processing unit, the terminal 1 includes a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), etc. The control unit issues execution commands to the processing unit, imaging unit, communication unit, input unit, output unit, and storage unit described later, and the processing unit performs data calculations and makes determinations on calculation results.

[0040] The imaging unit includes a device for capturing images such as videos and / or still images. There is no particular limitation as long as a color image can be captured.

[0041] As a communication unit, the terminal 1 includes a device for enabling communication with other terminals, devices, etc. The communication method may be wireless or wired.

[0042] As an input unit, the terminal 1 is assumed to have functions necessary for a user to operate the terminal 1. As examples for realizing input, it is possible to include a liquid crystal display for realizing a touch panel function, a keyboard, a mouse, a pen tablet, a hardware button on the device, a microphone for performing voice recognition, etc. The functions of the present invention are not particularly limited by the input method.

[0043] The terminal 1 shall be equipped with functions necessary for the user to operate the terminal 1 as an output unit. As examples of realizing output, forms such as display on a display unit 17 such as a liquid crystal display, a PC display, a projector, etc., and audio output are conceivable. The present invention is not particularly limited in function by the output method.

[0044] The terminal 1 shall be equipped with data storage using a hard disk, a semiconductor memory, a recording medium, a memory card, etc. as a storage unit. The storage destination of the data may be a cloud service, a database, or the like.

[0045] The control unit, in cooperation with the processing unit, realizes the acquisition unit 10, the detection unit 11, the tilt correction unit 12, the size correction unit 13, the recognition unit 14, the estimation unit 15, and the learning unit 16.

[0046] The above is the system configuration of the terminal 1 for recognizing the container number.

[0047] [Container Number Recognition Process] Based on FIG. 3, the container number recognition process executed by the terminal 1 for recognizing the container number will be described. FIG. 3 is a diagram showing a flowchart of the container number recognition process executed by the terminal 1 for recognizing the container number. As shown in FIG. 3, the container number recognition process is composed of steps S11 to S15 and corresponds to the above-described steps S1 to S5.

[0048] First, the acquisition unit 10 of the terminal 1 acquires an image including the entire vertically written container number displayed on the container (step S11). Specifically, the acquisition unit 10 acquires an image in which the container number displayed on the container photographed by the user using the imaging unit of the terminal 1 is shown. The image may be a color image, a moving image, or a still image.

[0049] Next, the detection unit 11 of the terminal 1 detects the area 100 of the entire container number in the acquired image, and detects each character 110 from the uppermost character to the lowermost character in the area (step S12). Specifically, as shown in FIG. 4, the detection unit 11 detects a rectangular area 100 including the entire vertically written container number shown in the image acquired in step S11, and detects the characters included in the detected area one by one from the uppermost to the lowermost, and encloses and extracts them in a rectangular shape so that each character 110 extends throughout, that is, extracts an image of a rectangular shape including each character 110. The method for detecting and extracting characters is not particularly limited. In the present embodiment, the area 100 is detected as a rectangle, but it may be detected in other shapes. Also, although each character 110 is enclosed in a rectangular shape, it may be enclosed in any other shape.

[0050] Also, in step S12, the estimation unit 15 of the terminal 1 may estimate the rectangular area 100 of the entire container number in the acquired image and estimate each character 110 from the uppermost character to the lowermost character in the rectangular area 100 based on the learned model 200 stored in the storage unit of the terminal 1. Specifically, the estimation unit 15 of the terminal 1 estimates and extracts the area 100 surrounded by a rectangle of the entire container number shown in the acquired image based on the learned model 200 created by machine learning in advance, and estimates an image of the city name area including each character 110 included in the extracted image. The type and method of machine learning are not particularly limited. The estimation method and extraction method of the area 100 of the entire container number and each character 110 are not particularly limited. Machine learning will be described later.

[0051] In step S12, the learning unit 16 of the terminal 1 may relearn, by machine learning, the feature amount specific to the detected or estimated area 100 of the entire container number and the feature amount specific to each detected or estimated character 110 based on the learned model 200 and the detection result or estimation result, and update the learned model 200. Note that this relearning by machine learning may be performed at any timing as long as it is after step S12, or may be performed by the terminal 1 receiving an input from the user. Machine learning will be described later.

[0052] Next, the inclination correction unit 12 of the terminal 1 determines the inclination from the coordinates of the uppermost character and the lowermost character, and corrects the inclination (step S13). Specifically, as shown in FIG. 5, the inclination correction unit 12 specifies the center C1 of the rectangular image including the uppermost character, the center C2 of the rectangular image including the lowermost character, and the center C0 of the character string extracted in step S12, and in the two-dimensional coordinate XY with the center C0 of the character string as the origin (0, 0), the coordinates (X1, Y1) of C1 of the uppermost character and the coordinates (X2, Y2) of C2 of the lowermost character are obtained. Then, by connecting C1 of the uppermost character and C2 of the lowermost character with a straight line and obtaining the inclination m of the straight line, a perpendicular line passing through the center C0 of the character string, that is, the origin (0, 0) is obtained. As a result, the coordinates of C1 of the uppermost character string become (0, Y1'), and the coordinates of the center C2 of the lowermost character become (0, Y2'), and as a result, the inclined vertically written character string is corrected to a straight vertically written character string.

[0053] Next, the size correction unit 13 of the terminal 1 calculates the average value of the sizes of the detected characters, and corrects the sizes of the characters to the average value (step S14). Specifically, as shown in FIG. 6, the size correction unit 13 measures the height of the rectangle surrounding each character in the vertically written character string corrected in step S13 in terms of the number of pixels, and calculates the average value of the measured rectangular heights of each character. Then, the rectangular image of each character is enlarged or reduced so that the height of each character becomes the average value.

[0054] Next, the recognition unit 14 of the terminal 1 combines the characters with corrected sizes side by side horizontally and recognizes the container number (step S15). Specifically, as shown in FIG. 7, the recognition unit 14 transposes the vertically arranged character string with corrected size in step S14 from rows to columns to make the character string horizontal, and then recognizes each character as character data. The output unit of the terminal 1 may output the recognized character data horizontally on the display unit 17. Note that the method for recognizing character data is not particularly limited.

[0055] The above is the container number recognition process.

[0056] [Learning process] Based on FIG. 8, the learning process executed by the terminal 1 for recognizing the container number will be described. FIG. 8 is a diagram showing a flowchart of the learning process executed by the terminal 1 for recognizing the container number.

[0057] The acquisition unit 10 of the terminal 1 acquires a learning image (step S31). Specifically, the acquisition unit 10 acquires an image in which the container number displayed on the photographed container is shown. The image may be a color image, a moving image, or a still image. Note that the learning image may be acquired from another terminal, computer, or device via data communication or the like.

[0058] The learning unit 16 of the terminal 1 analyzes the acquired learning image (step S32). Specifically, the learning unit 16 detects a rectangular area 100 including the entire vertically written container number shown in the acquired image, performs image analysis, and extracts feature amounts specific to the rectangular area 100. Further, the characters included in the detected rectangular area 100 are detected one by one from the top to the bottom, and image analysis is performed to extract feature amounts specific to each character 110. Note that the learning unit 16 may perform analysis of the acquired learning image by receiving an input from the user by the input unit of the terminal 1.

[0059] The learning unit 16 of the terminal 1 executes machine learning based on the analysis result (step S33). Specifically, the learning unit 16 executes supervised learning using the feature amounts extracted in step S32 as the teacher in machine learning. The feature amounts extracted in step S32 may be learned by deep learning that automatically defines and learns by a neural network having a multi-layer structure.

[0060] The learning unit 16 of the terminal 1 creates a learned model 200 for container number recognition based on the learning result (step S34). Specifically, the learning unit 16 creates a learned model 200 for container number recognition incorporating the feature amounts specific to the learned rectangular area 100 and the feature amounts specific to each character 110.

[0061] The learning unit 16 of the terminal 1 stores the created learned model 200 in its own storage unit (step S35).

[0062] The above is the learning process.

[0063] According to the terminal 1 for recognizing the container number, even if the container number is displayed vertically on the container, the container number can be read, so the operator does not need to handwrite the container number on paper or input it into the terminal. Therefore, similar to the case where the container number is displayed horizontally on the container, the location management of the container becomes easy.

[0064] Also, according to the terminal 1 for recognizing the container number, even if the container number is displayed vertically on the container, on the terminal, the container number is displayed horizontally. Therefore, it is possible to uniformly manage all the container numbers of the containers to be managed in horizontal writing.

[0065] Also, according to the terminal 1 for recognizing the container number, by learning the entire area 100 of the container number and each character 110 from the topmost character to the bottommost character in the area 100 to create the learned model 200, it is possible to improve the automatic detection accuracy of the entire area 100 of the container number and each character 110 in the area.

[0066] Furthermore, according to the terminal 1 for recognizing the container number, even when an image including the entire container number is captured at a distance by a fixed camera or the like, character recognition can be performed from the captured image based on the learning data, so more accurate character recognition becomes possible.

[0067] [Another aspect of this embodiment] [Outline of the system 2 for recognizing the container number] An overview of the system 2 for recognizing container numbers, which is another aspect of the present embodiment of the present invention, will be described with reference to FIG. 9. FIG. 9 is a diagram for explaining the overview of the system 2 for recognizing container numbers, which is another aspect of the present embodiment of the present invention. Note that the same reference numerals are given to the same functions and configurations as those in the above-described aspect, and the description thereof will be omitted. The difference between this aspect and the above-described aspect is that it is the system 2 for recognizing container numbers composed of the computer 3 and the imaging device 4.

[0068] As described above, the system 2 for recognizing container numbers is composed of the computer 3 and the imaging device 4.

[0069] The computer 3 is, for example, a computer such as a desktop personal computer, a notebook personal computer, or a server, a portable terminal such as a smartphone or a tablet terminal, or a wearable terminal such as a head-mounted display such as smart glasses or a smartwatch.

[0070] The computer 3 may be realized by, for example, a single terminal device, may be realized by a plurality of terminal devices, or may be realized by a virtual device such as a cloud computer.

[0071] The imaging device 4 is, for example, a camera that captures images such as moving images and / or still images, and is not particularly limited as long as it can capture color images.

[0072] The computer 3 and the imaging device 4 are connected so as to be capable of data communication via a public communication network or the like, and execute transmission and reception of necessary data and information.

[0073] Next, an overview of the processing executed by the system 2 for recognizing container numbers will be described. First, the computer 3 acquires an image including the entire vertically written container number displayed on the container (step S6). Specifically, the computer 3 acquires from the imaging device 4 an image in which the container number displayed on the container photographed by the user using the imaging device 4 is shown.

[0074] Next, the computer 3 detects the area of the entire container number in the acquired image, and detects each character from the topmost character to the bottommost character in the area (step S7). This step is the same process as step S2 of the above-described aspect, except that the computer 3 performs the process instead of the terminal 1.

[0075] Next, the computer 3 determines the inclination from the coordinates of the topmost character and the bottommost character, and corrects the inclination (step S8). This step is the same process as step S3 of the above-described aspect, except that the computer 3 performs the process instead of the terminal 1.

[0076] Next, the computer 3 calculates the average value of the sizes of the detected characters, and corrects the sizes of the characters with the average value (step S9). This step is the same process as step S4 of the above-described aspect, except that the computer 3 performs the process instead of the terminal 1.

[0077] Next, the computer 3 combines the characters with corrected sizes side by side to recognize the container number (step S10). This step is the same process as step S5 of the above-described aspect, except that the computer 3 performs the process instead of the terminal 1.

[0078] The above is an outline of the process executed by the system 2 for recognizing the container number.

[0079] [System Configuration of System 2 for Recognizing Container Number] Based on FIG. 10, the system configuration of the system 2 for recognizing the container number of the above-described aspect will be described.

[0080] The computer 3 is, for example, a computer such as a desktop personal computer, a notebook personal computer, or a server, a portable terminal such as a smartphone or a tablet terminal, a wearable terminal such as a head-mounted display such as smart glasses or a smartwatch, and the like.

[0081] The computer 3 may be implemented by, for example, a single terminal device, may be implemented by a plurality of terminal devices, or may be implemented by a virtual device such as a cloud computer.

[0082] The computer 3 includes, as a control unit and a processing unit, a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), etc. The control unit issues execution instructions to the processing unit, communication unit, input unit, output unit, and storage unit described later, and the processing unit performs data calculation and makes judgments on calculation results.

[0083] The computer 3 includes, as a communication unit, a device for enabling communication with other terminals, devices, etc. The communication method may be wireless or wired.

[0084] The imaging device 4 is, for example, a camera that captures images such as moving images and / or still images, and is not particularly limited as long as it can capture color images.

[0085] The computer 3 is assumed to include functions necessary for a user to operate the terminal 1 as an input unit. As an example for realizing input, it is possible to include a liquid crystal display that realizes a touch panel function, a keyboard, a mouse, a tablet, a hardware button on the device, a microphone for performing voice recognition, etc. The present invention is not particularly limited by the input method.

[0086] The computer 3 is assumed to include functions necessary for a user to operate the terminal 1 as an output unit. As an example for realizing output, forms such as display on a display unit 17 such as a liquid crystal display, a display of a PC, projection by a projector, etc., and voice output are conceivable. The present invention is not particularly limited by the output method.

[0087] The computer 3 has, as a storage unit, data storage by means of a hard disk, semiconductor memory, recording medium, memory card, etc. The data storage destination may be a cloud service, database, etc.

[0088] The control unit, in cooperation with the processing unit, realizes the acquisition unit 10, detection unit 11, tilt correction unit 12, size correction unit 13, recognition unit 14, estimation unit 15, and learning unit 16.

[0089] The imaging device 4 has, as a control unit, a CPU (Central Processing Unit), GPU (Graphics Processing Unit), RAM (Random Access Memory), ROM (Read Only Memory), etc. The control unit issues execution instructions to the processing unit, imaging unit, communication unit, input unit, output unit, and storage unit described later. The processing unit performs data calculations and makes determinations on calculation results, etc.

[0090] The imaging device 4 is assumed to have functions necessary for shooting images such as videos and / or still images, etc. If a color image can be shot, it is not particularly limited.

[0091] The imaging device 4 has, as a communication unit, a device for enabling communication with other terminals, devices, etc. The communication method may be wireless or wired.

[0092] The imaging device 4 is assumed to have functions necessary for a user to operate the imaging device 4 as an input unit. As an example for realizing input, it is possible to have a liquid crystal display that realizes a touch panel function, etc. The functions of the present invention are not particularly limited by the input method.

[0093] The imaging device 4 is assumed to have functions necessary for a user to operate the imaging device 4 as an output unit. As an example for realizing output, forms such as a liquid crystal display, etc. are conceivable. The functions of the present invention are not particularly limited by the output method.

[0094] The imaging device 4 includes, as a storage unit, data storage using a hard disk, semiconductor memory, recording medium, memory card, or the like. The storage destination of the data may be a cloud service, database, or the like.

[0095] The computer 3 and the imaging device 4 are connected to be capable of data communication via a public communication network or the like, and perform transmission and reception of necessary data and information.

[0096] The above is the system configuration of the system 2 for recognizing the container number.

[0097] [Container Number Recognition Process] The container number recognition process executed by the system 2 for recognizing the container number will be described. FIG. 11 is a diagram showing a flowchart of the container number recognition process executed by the system 2 for recognizing the container number. As shown in FIG. 11, the container number recognition process is composed of steps S21 to S25 and corresponds to steps S6 to S10 described above.

[0098] First, the acquisition unit 10 of the computer 3 acquires an image including the entire vertically written container number displayed on the container (step S21). Specifically, the acquisition unit 10 acquires, from the imaging device 4, an image in which the container number displayed on the container photographed by the user using the imaging device 4 is shown. The difference from the above-described one aspect is that the computer 3 acquires an image in which the container number displayed on the container photographed by the user using the imaging device 4 is shown. The acquisition unit 10 may acquire the image photographed by the imaging device 4 via a communication network by the communication unit of the computer 3, or may acquire it by being stored in the storage unit by an external medium.

[0099] Next, the detection unit 11 of the computer 3 detects the area of the entire container number in the acquired image, and detects each character from the topmost character to the bottommost character in the area (step S22). This step is the same process as step S12 of the above-described one aspect, but is different in that the computer 3 performs the process instead of the terminal 1.

[0100] Next, the inclination correction unit 12 of the computer 3 determines the inclination from the coordinates of the uppermost and lowermost characters and corrects the inclination (step S23). This step is the same process as step S13 in the above-described aspect, but is different in that the computer 3 performs the process instead of the terminal 1.

[0101] Next, the size correction unit 13 of the computer 3 calculates the average value of the sizes of the detected characters and corrects the sizes of the characters to the average value (step S24). This step is the same process as step S14 in the above-described aspect, but is different in that the computer 3 performs the process instead of the terminal 1.

[0102] Next, the recognition unit 14 of the computer 3 combines the characters with corrected sizes horizontally and recognizes the container number (step S25). This step is the same process as step S15 in the above-described aspect, but is different in that the computer 3 performs the process instead of the terminal 1.

[0103] The above is the container number recognition process.

[0104] [Learning process] Regarding the learning process executed by the system 2 for recognizing the container number, the description is omitted because it is different from the above-described aspect only in that the computer 3 acquires an image in which the entire image of the serial number of the tire photographed by the user using the imaging device 4 is shown.

[0105] According to the system 2 for recognizing the container number, as in the above-described aspect, even if the container number is displayed vertically on the container, the container number can be read. Therefore, the operator does not need to handwrite the container number on paper or input it into the terminal. Thus, similar to the case where the container number is displayed horizontally on the container, the location management of the container becomes easy.

[0106] Also, according to the system 2 for recognizing the container number, similar to the above-described aspect, even if the container number is displayed vertically on the container, on the terminal, the container number is displayed horizontally. Therefore, it is possible to uniformly manage all the container numbers of the containers to be managed in horizontal writing.

[0107] Also, according to the system 2 for recognizing the container number, similar to the above-described aspect, by learning the entire area 100 of the container number and each character 110 from the topmost character to the bottommost character in the area 100 to create the learned model 200, it is possible to improve the automatic detection accuracy of the entire area 100 of the container number and each character 110 in the area.

[0108] Also, according to the system 2 for recognizing the container number, similar to the above-described aspect, even when an image including the entire container number is captured at a distance by a stationary camera or the like, character recognition can be performed from the captured image by the learning data, so that more accurate character recognition is possible.

[0109] Furthermore, instead of the terminal 1 for recognizing the container number, different from the above-described aspect, as a system for recognizing the container number, it can be realized by a computer 3 such as a desktop personal computer, a notebook personal computer, a server, etc., and an imaging device 4 such as a camera for capturing images such as color videos and / or still images. Therefore, by installing the imaging device 4 at a certain location, when a container is placed or passes through that location, the computer 3 can acquire an image of the area 100 including the container number captured by remote operation and recognize the container number, which can save the user's time and effort of going to a certain location.

[0110] The above-described means and functions are realized by a computer (including a CPU, an information processing device, and various terminals) reading and executing a predetermined program. The program is provided, for example, in a form provided via a network from one or more terminals (in the form of cloud services, SaaS: Software as a Service). Also, the program is provided, for example, in a form recorded on a computer-readable recording medium. In this case, the computer reads the program from the recording medium, transfers it to and records it in an internal recording device or an external recording device, and then executes it. Further, the program may be pre-recorded in a recording device (recording medium) such as a magnetic disk, an optical disk, or a magneto-optical disk, and provided to a terminal via a communication line from the recording device.

[0111] As described above, the embodiments of the present invention have been explained, but the present invention is not limited to these above-described embodiments. Also, the effects described in the embodiments of the present invention are merely an enumeration of the most preferable effects resulting from the present invention, and the effects according to the present invention are not limited to those described in the embodiments of the present invention.

Explanation of Reference Numerals

[0112] 1 Terminal, 2 System, 3 Computer, 4 Imaging Device, 10 Acquisition Unit, 11 Detection Unit, 12 Tilt Correction Unit, 13 Size Correction Unit, 14 Recognition Unit, 15 Estimation Unit, 16 Learning Unit, 17 Display Unit, 100 Region, 110 Each Character, 200 Learned Model

Claims

1. A terminal for recognizing a container number, comprising: an acquisition unit that acquires an image including the entire vertically-written container number displayed on the container; a detection unit that detects the region of the entire container number in the acquired image and detects each character from the uppermost character to the lowermost character in the region; an inclination correction unit that determines an inclination from the coordinates of the uppermost character and the lowermost character and corrects the inclination; a size correction unit that calculates an average value of the sizes of the detected characters and corrects the sizes of the characters to the average value; a recognition unit that combines the characters with the corrected size side by side and recognizes the characters as the container number; A terminal comprising the above.

2. The terminal according to claim 1, further comprising a display unit that displays the recognized container number in horizontal writing.

3. The terminal according to claim 1, further comprising a learning unit that learns the region of the entire container number and each character in the region to create a learned model.

4. The terminal according to claim 3, further comprising an estimation unit that estimates the region of the entire container number in the acquired image based on the learned model and estimates each character from the uppermost character to the lowermost character in the region.

5. A system for recognizing a container number, comprising: an acquisition unit that acquires an image including the entire vertically-written container number displayed on the container; a detection unit that detects the region of the entire container number in the acquired image and detects each character from the uppermost character to the lowermost character in the region; an inclination correction unit that determines an inclination from the coordinates of the uppermost character and the lowermost character and corrects the inclination; a size correction unit that calculates an average value of the sizes of the detected characters and corrects the sizes of the characters to the average value; a recognition unit that combines the characters with the corrected size side by side and recognizes the characters as the container number; A system comprising the above.

6. A method for recognizing a container number executed by a computer system, comprising: a step of acquiring an image including the entire vertically-written container number displayed on the container; a step of detecting the region of the entire container number in the acquired image and detecting each character from the uppermost character to the lowermost character in the region; a step of determining an inclination from the coordinates of the uppermost character and the lowermost character and correcting the inclination; Calculating an average value of the sizes of the detected respective characters, and correcting the sizes of the respective characters to the average value; Combining the respective characters with the corrected sizes horizontally, and recognizing the respective characters as the container numbers; A method comprising the above steps.

7. Causing a computer to acquire an image including the entire vertically-written container number displayed on a container; detect a region of the entire container number in the acquired image, and detect each character from the uppermost character to the lowermost character in the region; determine an inclination from coordinates of the uppermost character and the lowermost character, and correct the inclination; calculate an average value of the sizes of the detected respective characters, and correct the sizes of the respective characters to the average value; combine the respective characters with the corrected sizes horizontally, and recognize the respective characters as container numbers; A computer-readable program for causing the above steps to be executed.

Images