Detection device and detection method

Abstract

To detect a trailer ID or a container ID written on a trailer.SOLUTION: A detection device includes: a trailer area detection part for detecting a trailer area being an area where a trailer is photographed from a picked-up image obtained by photographing at least a part of a vehicle to which a tractor and the trailer are connected; and an ID area output part for detecting an ID area being an area estimated to be an area with a trailer ID or a container ID written from the trailer area, and outputting information showing the ID area to a character recognition part for recognizing the trailer ID or the container ID by character recognition to the ID area.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present disclosure relates to a detection device and a detection method. [Background technology]

[0002] Patent Document 1 discloses a license plate recognition device that detects multiple quadrangles of license plate area candidates from an input image and performs character recognition on the character areas included in the license plate area candidates. The license plate recognition device selects a license plate area candidate to output from the multiple detected license plate area candidates based on the character recognition results and quadrangle information for each license plate area candidate. The license plate recognition device then outputs information about the selected license plate area candidate. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2008-217347 Summary of the Invention [Problem to be solved by the invention]

[0004] In recent years, logistics using trailers has become popular. Trailers are towed by tractors and enter and leave yards (e.g., warehouse facilities). An identification (ID) (hereinafter referred to as a trailer ID) for identifying the trailer is written on the trailer. Therefore, the use of trailer IDs to identify trailers entering and leaving yards has been studied. However, trailers may have various characters, patterns, and designs in addition to the trailer ID written on them, making it difficult to accurately detect the trailer ID.

[0005] The present disclosure aims to accurately detect a trailer ID or container ID written on a trailer. [Means for solving the problem]

[0006] One aspect of the present disclosure provides a detection device including: a trailer area detection unit that detects a trailer area, which is an area in which the trailer is photographed, from an image capturing at least a portion of a vehicle in which a tractor and a trailer are coupled; and an ID area detection unit that detects an ID area, which is an area in which a trailer ID or a container ID is presumed to be written, from the trailer area, and outputs information indicating the ID area to a character recognition unit that recognizes the trailer ID or the container ID through character recognition of the ID area.

[0007] One aspect of the present disclosure provides a detection method that detects a trailer area, which is the area where the trailer is photographed, from an image capturing at least a portion of a vehicle having a tractor and a trailer connected thereto, detects an ID area, which is an area where a trailer ID or a container ID is presumed to be written, from the trailer area, and outputs information indicating the ID area to a character recognition unit that recognizes the trailer ID or the container ID through character recognition of the ID area.

[0008] These comprehensive or specific aspects may be realized as a system, an apparatus, a method, an integrated circuit, a computer program, or a recording medium, or may be realized as any combination of a system, an apparatus, a method, an integrated circuit, a computer program, and a recording medium. [Effects of the Invention]

[0009] According to the present disclosure, the trailer ID or container ID written on the trailer can be detected with high accuracy. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a diagram showing an example of a yard according to the first embodiment. [Figure 2] FIG. 1 is a block diagram illustrating an example of a management system according to a first embodiment. [Figure 3]FIG. 1 is a diagram for explaining a first method for detecting a tractor area and a trailer area according to the first embodiment; [Figure 4] FIG. 10 is a diagram for explaining a second method for detecting a tractor area and a trailer area according to the first embodiment. [Figure 5] FIG. 1 is a diagram for explaining a method for detecting a trailer ID area according to the first embodiment. [Figure 6] FIG. 10 is a block diagram illustrating an example of a management system according to a second embodiment. [Figure 7] FIG. 1 is a first diagram for explaining resizing of a captured image according to a second embodiment; [Figure 8] FIG. 2 is a second diagram illustrating resizing of a captured image according to the second embodiment; [Figure 9] Graph for explaining an example of a resize magnification of a captured image according to the second embodiment. [Figure 10] FIG. 1 is a block diagram illustrating an example of a hardware configuration of a detection device according to the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments of the present disclosure will be described in detail with appropriate reference to the drawings. However, more detailed description than necessary may be omitted. For example, detailed descriptions of already well-known matters and redundant descriptions of substantially identical configurations may be omitted. This is to avoid unnecessary redundancy in the following description and to facilitate understanding by those skilled in the art. Note that the accompanying drawings and the following description are provided to enable those skilled in the art to fully understand the present disclosure and are not intended to limit the subject matter described in the claims. Furthermore, the function of one configuration shown in the present embodiment may be realized by two or more physical configurations, or the functions of two or more configurations may be realized by, for example, one physical configuration.

[0012] (Embodiment 1) FIG. 1 is a diagram showing an example of a yard according to the first embodiment.

[0013] As shown in FIG. 1, a tractor 11 towing a trailer 12 enters and exits a yard 1. The tractor 11 may be interpreted as a towing vehicle. The tractor 11 and trailer 12 enter the yard 1 through an entrance area 2 and move to a parking area 4. The tractor 11 may then detach (drop off) the trailer 12 in the parking area 4. The tractor 11 may also move to another parking area 4 where the trailer 12 is parked. The tractor 11 may then couple to the trailer 12 parked in the other parking area 4 and exit through an exit area 3 while towing the trailer 12.

[0014] Here, the trailer 12 refers to the entire object towed by the tractor 11. Generally, a vehicle towed by the tractor 11 is called a "trailer." The trailer 12 may be configured as a combination of a container that stores cargo but does not have a mechanism for movement, such as wheels, and a dolly on which the container is loaded. The trailer 12 may also be configured as a vehicle in which a portion for storing cargo and a mechanism for movement are integrated. Note that when the tractor 11 tows a combination of a container and a dolly, only the dolly may be referred to as a "trailer." However, for the sake of convenience, the present specification will refer to the entire object towed by the tractor 11, i.e., both the combination of the dolly and the container, or the vehicle equipped with a portion for storing cargo and a mechanism for movement, as a "trailer." The present specification will also sometimes refer to the coupled tractor 11 and trailer 12 collectively as simply a vehicle 10.

[0015] An imaging device 21 is installed at a position where it can capture images of the entrance area 2 and the exit area 3 of the yard 1.

[0016] Generally, an ID (hereinafter referred to as a trailer ID) for identifying the trailer 12 is written on the trailer 12. A management system 20 (see FIG. 2) according to the first embodiment manages the entry and exit of trailers 12 into and out of the yard 1 by detecting and managing the trailer ID written on the trailer 12 using captured images generated by an imaging device 21 capturing images of the tractor 11 and trailer 12 entering and leaving the yard 1. This makes it possible to efficiently manage the large number of trailers 12 entering and leaving the yard 1. The management system 20 will be described in detail below.

[0017] FIG. 2 is a block diagram showing an example of the management system 20 according to the first embodiment.

[0018] The management system 20 includes at least one imaging device 21, a detection device 30, and a management device 22. The imaging device 21, the detection device 30, and the management device 22 can transmit and receive information to and from each other via a predetermined communication network 23 (see FIG. 10).

[0019] The imaging device 21 is installed in a position where it can capture an image of the entrance area 2 and / or exit area 3 of the yard 1. The imaging device 21 captures an image of the tractor 11 and trailer 12 (vehicle 10) passing through the entrance area 2 and / or exit area 3, and generates a captured image 100 (see FIG. 3, etc.). Note that the imaging device 21 may be installed at any location in the yard 1, not limited to the entrance area 2 and / or exit area 3 of the yard 1.

[0020] The detection device 30 is a device that detects a trailer ID using a captured image 100 captured by the imaging device 21. For example, the detection device 30 detects a trailer ID from the captured image 100 captured by the imaging device 21 in the entrance area 2. The detection device 30 also detects a trailer ID from the captured image 100 captured by the imaging device 21 in the exit area 3.

[0021] The management device 22 is a device that manages the trailer IDs detected by the detection device 30. For example, the management device 22 receives and manages the trailer IDs detected from the captured image 100 of the entrance area 2 and the trailer IDs detected from the captured image 100 of the exit area 3 from the detection device 30. This allows the management device 22 to manage the trailers 12 that have entered the yard 1 and the trailers 12 that have left the yard 1.

[0022] The detection device 30 has, as its functions, a vehicle recognition unit 31, a trailer ID area detection unit 32, and a character recognition unit 33. These functions may be realized by a processor 1001 (see FIG. 10) of the detection device 30 reading and executing a program from a memory 1002 (see FIG. 10).

[0023] The vehicle recognition unit 31 recognizes the vehicle 10 (tractor 11 and trailer 12) by using a known image recognition technique from the captured image 100 received from the imaging device 21. The vehicle recognition unit 31 includes a trailer area detection unit 41 and a tractor area detection unit 42.

[0024] Fig. 3 is a diagram for explaining a first method for detecting a tractor area and a trailer area according to the first embodiment. Fig. 4 is a diagram for explaining a second method for detecting a tractor area and a trailer area according to the first embodiment.

[0025] 3, the trailer area detection unit 41 uses a known image recognition technique to detect an area in which the entire trailer 12 is captured from the captured image 100 as the trailer area 101A. For example, the trailer area detection unit 41 uses a learning model (e.g., a convolutional neural network (CNN)) trained by deep learning using images in which a large number of trailers 12 are captured to detect an area in which the entire trailer 12 is captured from the captured image 100 as the trailer area 101A.

[0026] 4, the trailer area detection unit 41 may detect an area in which the front or rear of the trailer 12 is captured from the captured image 100 as the trailer area 101, 101B. For example, the trailer area detection unit 41 uses a learning model (e.g., CNN) trained by deep learning using a large number of images in which the front or rear of the trailer 12 is captured to detect an area in which the front or rear of the trailer 12 is captured from the captured image 100 as the trailer area 101B. This is because the trailer ID is often written on the front or rear of the trailer 12.

[0027] The trailer area detection unit 41 may detect both the trailer area 101A in which the entire trailer 12 is imaged as shown in Fig. 3 and the trailer area 101B in which the front or rear of the trailer 12 is imaged as shown in Fig. 4. In this specification, the trailer area 101A in which the entire trailer 12 is imaged and the trailer area 101B in which the front or rear of the trailer 12 is imaged may be referred to as the trailer area 101 without distinction.

[0028] 3 or 4, the tractor area detection unit 42 uses a known image recognition technique to detect an area in which the tractor 11 is imaged from the captured image 100 as the tractor area 102. For example, the tractor area detection unit 42 uses a learning model (e.g., CNN) that is trained by deep learning using images in which a large number of tractors 11 are imaged to detect an area in which the tractor 11 is imaged from the captured image 100 as the tractor area 102.

[0029] FIG. 5 is a diagram illustrating a method for detecting a trailer ID area according to the first embodiment.

[0030] The trailer ID area detection unit 32 detects one or more areas where a trailer ID is presumed to be written from within the trailer area 101 of the captured image 100 detected by the trailer area detection unit 41, as the trailer ID area 110. For example, the trailer ID area detection unit 32 detects areas where a trailer ID is presumed to be written from within the trailer area 101 of the captured image 100, as the trailer ID area 110, using a learning model (e.g., CNN) trained by deep learning using images in which a large number of trailer IDs are captured.

[0031] Alternatively, the trailer ID area detection unit 32 performs optical character recognition (OCR) processing on the trailer area 101 of the captured image 100, and detects an area in which a character string that satisfies a predetermined condition related to a trailer ID is written as the trailer ID area 110. The predetermined condition related to a trailer ID is, for example, characteristics of a character string that is likely to be a trailer ID. Specifically, the character string is composed of alphabets and numbers, is written vertically, is similar to a known trailer ID, etc.

[0032] The trailer ID area detection unit 32 may detect the trailer ID directly from the captured image 100 without being limited to the trailer area 101 .

[0033] The character recognition unit 33 performs OCR processing on the trailer ID area 110 of the captured image 100 detected by the trailer ID area detection unit 32 to recognize a character string. Then, the character recognition unit 33 outputs the recognized character string as a trailer ID. If there are multiple trailer ID areas 110, the character recognition unit 33 may perform OCR processing on each trailer ID area 110.

[0034] The character recognition unit 33 may determine whether the recognized character string satisfies a predetermined condition for the trailer ID. The character recognition unit 33 may output, as the trailer ID, a character string that is determined to satisfy the predetermined condition for the trailer ID, and may not output, as the trailer ID, a character string that is determined not to satisfy the predetermined condition for the trailer ID.

[0035] The trailer ID output from the character recognition unit 33 is transmitted to, for example, the management device 22. This allows the management device 22 to manage the trailer IDs of the trailers 12 that have entered and exited the yard 1.

[0036] Furthermore, the trailer ID output from the character recognition unit 33 may be displayed on the display device 1005 (see FIG. 10) together with the captured image 100 from which the trailer ID was detected.

[0037] According to this embodiment, as shown in Fig. 3, the detection device 30 can detect the trailer area 101A using the trailer area detection unit 41 and detect the trailer ID within the trailer area 101A. This can prevent, for example, a character string written on the tractor 11 from being erroneously detected as the trailer ID.

[0038] 4, the detection device 30 can detect the trailer area 101B by using the trailer area detection unit 41, which is limited to the front or rear of the trailer 12, and can detect the trailer ID within the trailer area 101B. This can prevent, for example, a character string, logo, or illustration written on the side of the trailer 12 from being erroneously detected as the trailer ID.

[0039] Therefore, according to this embodiment, the detection device 30 can detect the trailer ID from the captured image 100 captured by the imaging device 21 with higher accuracy.

[0040] (Embodiment 2) In the second embodiment, the same reference numerals are used for the components already described in the first embodiment, and the description thereof may be omitted.

[0041] FIG. 6 is a block diagram showing an example of a management system 20 according to the second embodiment.

[0042] The management system 20 includes at least one imaging device 21, a detection device 30, and a management device 22. The imaging device 21 and the management device 22 are the same as those in the first embodiment.

[0043] The detection device 30 according to the second embodiment has, as its functions, a vehicle recognition unit 31, a resizing unit 34, a trailer ID area detection unit 32, and a character recognition unit 33. As in the first embodiment, the vehicle recognition unit 31 includes a trailer area detection unit 41 and a tractor area detection unit 42. Moreover, the trailer area detection unit 41 and the tractor area detection unit 42 are the same as in the first embodiment.

[0044] Fig. 7 is a first diagram for explaining resizing of a captured image according to Embodiment 2. Fig. 8 is a second diagram for explaining resizing of a captured image according to Embodiment 2. Fig. 9 is a graph for explaining an example of a resizing magnification of a captured image according to Embodiment 2. In graph 300 of Fig. 9, the horizontal axis represents the height H of the trailer area 101, and the vertical axis represents the resizing magnification.

[0045] For example, when the learning model is CNN, an object whose size corresponds to the convolution region has higher detection accuracy. Furthermore, the larger the convolution region, the longer the detection processing time. Therefore, when the trailer ID region detection unit 32 described in the first embodiment detects the trailer ID region 110 from the trailer region 101 using CNN, the detection accuracy of the trailer ID region 110 improves and the detection processing time shortens when the size of the trailer region 101 is constant. Therefore, the detection device 30 according to the second embodiment further includes a resizing unit 34, which appropriately resizes the captured image 100 (i.e., appropriately resizes the trailer region 101), before the trailer ID region detection unit 32.

[0046] The resizing unit 34 resizes the captured image 100 in accordance with the size of the trailer area 101 (for example, the height H of the trailer area 101) and generates a resized image 200.

[0047] That is, the resizing unit 34 determines the resizing ratio based on the height H of the trailer area 101. This is because using the trailer height allows for determining a more appropriate resizing ratio than using other information. The reason for this is explained below. Generally, legal restrictions are imposed on the height of trailers, and many trailers adopt a height close to the legal limit in order to maximize payload. As a result, it is highly likely that the height of actual trailers is approximately constant. Therefore, by resizing based on the height H of the trailer area 101 in the captured image 100, the size of the trailer included in the resized trailer area 101 can be made approximately constant. This not only makes the size of the trailer area 101 constant, but also makes the size of the trailer in which the trailer ID area 110 can be detected constant, thereby improving the detection accuracy of the trailer ID area 110 and shortening the detection processing time.

[0048] Although legal restrictions are imposed on the length and width of trailers, as with height, it is more advantageous to use height rather than these information in order to calculate an appropriate resizing ratio. This is due to the following reasons: Because trailers are generally very long, the front or back of the trailer may be cut off in the captured image 100, and accurate information about the trailer's overall length may not be obtained from the captured image 100. Furthermore, the front of the trailer may be hidden by a tractor, and the back may be hidden by another vehicle following it, so accurate information about the trailer's overall width may not be obtained from the captured image 100. Therefore, if the resizing ratio is determined based on the trailer's length or width, there is a risk that the resizing ratio will be determined based on inaccurate information.

[0049] Furthermore, due to constraints such as yard facilities and roads, the locations where the imaging device 21 can be installed and the angles at which it can capture images may be limited. Therefore, the direction from which the imaging device 21 is installed—front, rear, or side—from which the vehicle is captured may vary depending on the yard or road. In this case, if information that is difficult to obtain from the front but easy to obtain from the side, such as the length of the trailer, is used to determine the resizing magnification, whether an appropriate resizing magnification can be obtained will vary depending on the installation conditions of the imaging device 21. Therefore, it is necessary to change the information used to determine the resizing magnification for each installation location or imaging angle of the imaging device 21, which complicates processing. Furthermore, since the resizing magnification is determined based on different criteria depending on the conditions of the imaging device 21, there is a risk of instability in accuracy. On the other hand, trailer height information can be reliably obtained regardless of the direction from which the vehicle is captured. Therefore, by using the trailer height to determine the resizing magnification, the resizing magnification can be determined based on the same criteria regardless of the installation location of the imaging device 21 or the angle at which it can capture images. This simplifies processing and stabilizes accuracy.

[0050] For example, as shown in graph 300 in Fig. 9, when the height H of the trailer area 101 is less than a predetermined first threshold (e.g., 900 pix), the resizing unit 34 may set the resizing magnification to 1.0. In other words, when the height H of the trailer area 101 is less than the predetermined first threshold, the resizing unit 34 does not need to resize the captured image 100, as shown in Fig. 7.

[0051] For example, as shown in graph 300 in Fig. 9, when the height H of the trailer area 101 is equal to or greater than a predetermined first threshold, the resizing unit 34 may reduce the resizing factor in proportion to the height H. In other words, when the height H of the trailer area 101 is equal to or greater than the predetermined first threshold, the resizing unit 34 may reduce the captured image 100 so that the inside of the trailer area 101 of the captured image 100 has a predetermined size, as shown in Fig. 8.

[0052] However, a lower limit value of the resize magnification may be set in advance as shown in graph 300 of Fig. 9. For example, when the height H of the trailer area 101 is equal to or greater than a predetermined second threshold value that is greater than the first threshold value, the resize unit 34 may set the resize magnification to a lower limit value (for example, 0.1 times).

[0053] The resizing unit 34 also resizes the trailer area 101 in accordance with the resizing of the captured image 100. As a result, for example, as shown in Fig. 8, the trailer area 201 is set in the area in the resized image 200 where the front or rear of the trailer 12 is captured. Then, the trailer area 201 of the resized image 200 becomes a predetermined size.

[0054] The trailer ID area detection unit 32 uses, for example, CNN to detect one or more areas in which a trailer ID is presumed to be written from within the trailer area 201 of the resized image 200, as the trailer ID area 110. Here, since the trailer area 201 has been resized to a predetermined size, the trailer ID area detection unit 32 can accurately detect, in a short time, the trailer ID area 110 in which a trailer ID is presumed to be written from within the trailer area 201 using CNN of a predetermined convolution area.

[0055] As in the first embodiment, the character recognition unit 33 performs OCR processing on the trailer ID area 110 of the resized image 200 to recognize a character string, and outputs the recognized character string as a trailer ID.

[0056] According to this embodiment, the resizing unit 34 resizes (e.g., reduces) the captured image 100 to a size suitable for the size of a predetermined convolution area in the trailer ID area detection unit 32, and generates a resized image 200. This allows the trailer ID area detection unit 32 to detect the trailer ID area 110 by applying the predetermined convolution area to the resized image 200, and therefore the trailer ID area 110 can be detected accurately in a short time.

[0057] (Hardware configuration) FIG. 9 is a block diagram showing an example of the hardware configuration of the detection device 30 according to the present disclosure.

[0058] The detection device 30 includes a processor 1001 , a memory 1002 , a storage 1003 , an input device 1004 , a display device 1005 , and a communication device 1006 .

[0059] The processor 1001 is a device that executes a computer program stored in the memory 1002 and realizes the above-described functions of the detection device 30. The processor 1001 may be interpreted as a central processing unit (CPU), a controller, a control unit, or a control device. The processor 1001 may also include a graphics processing unit (GPU) and / or a neural processing unit (NPU). The processing using the above-described learning model may be executed by the GPU and / or the NPU.

[0060] The memory 1002 is configured as a volatile storage medium and / or a non-volatile storage medium, and is a device for storing computer programs and data handled by the detection device 30.

[0061] The storage 1003 is configured as a non-volatile storage medium, and is a device that stores computer programs and data handled by the detection device 30. Examples of the storage 1003 include a hard disk drive (HDD), a solid state drive (SSD), or a flash memory.

[0062] The input device 1004 is a device that accepts data from a user to be input to the processor 1001. Examples of the input device 1004 include a keyboard, a mouse, a touchpad, and a microphone.

[0063] The display device 1005 is a device that displays data generated by the processor 1001. Examples of the display device 1005 include a liquid crystal display and an organic EL display. For example, the display device 1005 displays the captured image 100 and the trailer ID detected from the captured image 100.

[0064] The communication device 1006 is a device that transmits and receives data to and from the imaging device 21, the management device 22, etc. via a communication network 23. Examples of the communication network 23 include a wired local area network (LAN), a wireless LAN, a mobile communication network, the Internet, etc.

[0065] (Other variations) In the above-described embodiments, the tractor region detection unit 42 detects the tractor region 102. However, as described above, when the trailer ID region 110 is detected using the trailer region 101, it is not necessary to use the tractor region 102. Therefore, the tractor region detection unit 42 may be omitted in each embodiment. When detecting the tractor region 102, the accuracy of detecting the trailer region 101 may be improved by determining that the portion included in the tractor region 102 is not part of the trailer region 101. Even if character recognition is successful, if the region in which the character string is detected overlaps the trailer region 101 and the tractor region 102, the result of the character recognition may be discarded as a false positive. This is because tractors may also have numeric strings or character strings written on them. If the trailer region 101 includes an area in which such numeric strings or character strings are written, the numeric strings or character strings of the tractor may be mistakenly recognized as a trailer ID. Note that this problem would not occur if the tractor area and trailer area were configured as areas that precisely extract the ranges in which the tractor and trailer are present, rather than as rectangular areas. However, accurately and precisely extracting the tractor area and trailer area increases the processing load and the possibility of erroneous detection of each area. For this reason, rectangular areas are used as these areas in the above-described embodiments and this modification.

[0066] In each of the above-described embodiments, the trailer ID area detected by the trailer ID area detection unit 32 is output to the character recognition unit 33. However, the information output by the trailer ID area detection unit 32 is not limited to the trailer ID area. For example, the trailer ID area detection unit 32 may output the coordinates of the trailer ID area, and the character recognition unit 33 may acquire the captured image 100 itself and extract the portion corresponding to the coordinates output by the trailer ID area detection unit 32 to determine the trailer ID area. In other words, the format of the content output by the trailer ID area detection unit 32 is not important as long as it is information indicating the extracted area.

[0067] In the above-described embodiments, the character recognition unit 33 is described as being included in the detection device 30. However, the character recognition unit 33 may be provided external to the detection device 30. In this case, the character recognition unit 33 may be included in a server shared by multiple detection devices or multiple yards. Because trailer IDs are determined by the trailer management company or the like, trailer IDs written in the same format may be used in different locations. For this reason, it may be possible to stabilize the accuracy of character recognition for the entire system by consolidating the character recognition function in a server or the like, rather than having the detection device 30 or the yard perform character recognition independently.

[0068] In the above-described embodiments, the configuration for detecting the trailer ID area 110 has been described. However, the concept of the above-described embodiments can also be applied to applications for detecting other IDs written on the trailer. For example, the configuration may be used to detect the container ID area, which is an area where a container ID for identifying a container is likely to be written. Here, the trailer ID is an ID defined in a unique format by each trailer manager, while the container ID is a code defined in international standards such as ISO 6346 for managing containers. Furthermore, the configuration may be used to detect an area where an ID for identifying a trailer or container is written, without distinguishing between the trailer ID and the container ID.

[0069] Summary of the Disclosure Based on the above description of the present disclosure, the following techniques are disclosed.

[0070] <Technology 1> The detection device (30) according to the present disclosure includes a trailer area detection unit (41) that detects a trailer area (101) in which the trailer is photographed from an image (100) capturing at least a portion of a vehicle (10) having a tractor (11) and a trailer (12) coupled thereto, and an ID area detection unit (e.g., trailer ID area detection unit 32) that detects an ID area (e.g., trailer ID area 110) in which a trailer ID or a container ID is presumed to be written from the trailer area and outputs information indicating the ID area to a character recognition unit (33) that recognizes the trailer ID or the container ID by character recognition of the ID area. In this way, by detecting the trailer area from the captured image and then detecting the ID area from the trailer area, it is possible to prevent, for example, a character string written on a tractor from being mistakenly detected as a trailer ID or container ID, thereby improving the detection accuracy of the trailer ID or container ID.

[0071] <Technology 2> In the detection device described in Technique 1, the trailer area detection unit detects an area in which the front or rear of the trailer is photographed from the captured image as the trailer area. In this way, by defining the area where the front or rear of the trailer is photographed as the trailer area, it is possible to prevent, for example, a character string written on the side of the trailer from being erroneously detected as the trailer ID, thereby improving the detection accuracy of the trailer ID.

[0072] <Technology 3> The detection device according to Technology 1 or 2 further includes a resizing unit (34) that resizes the captured image based on the size of the trailer area to generate a resized image, and the trailer ID area detection unit detects the trailer ID area from the trailer area of ​​the resized image. This allows the trailer ID area detection unit to detect the trailer ID area from the trailer area resized to an appropriate size, thereby reducing the processing time for detecting the trailer ID area and improving the detection accuracy of the trailer ID area.

[0073] <Technology 4> In the detection device described in Technique 3, the resizing unit resizes the captured image based on the height of the trailer area. This allows the trailer ID area detection unit to detect the trailer ID area from the trailer area resized to an appropriate size, thereby reducing the processing time for detecting the trailer ID area and improving the detection accuracy of the trailer ID area.

[0074] <Technology 5> In the detection device described in Technology 4, the resizing unit reduces the captured image when the height of the trailer area is equal to or greater than a predetermined threshold, and does not reduce the captured image when the height of the trailer area is less than the threshold. As a result, when the captured image is relatively large, the trailer ID area can be detected from the trailer area reduced to an appropriate size, thereby shortening the processing time required to detect the trailer ID area.

[0075] <Technology 6> In the detection device described in Technique 1, the ID region detection unit excludes regions other than the trailer region from the detection target of the ID region. This prevents, for example, a character string written on a tractor from being mistakenly detected as a trailer ID or container ID.

[0076] <Technology 7> The detection method according to the present disclosure detects a trailer area (101) in which the trailer is photographed from an image (100) capturing at least a portion of a vehicle (10) having a tractor (11) and a trailer (12) coupled thereto, and detects an ID area (e.g., trailer ID area 110) in which a trailer ID or a container ID is presumed to be written from the trailer area, and outputs information indicating the ID area to a character recognition unit (33) that recognizes the trailer ID or the container ID by character recognition of the ID area. In this way, by detecting the trailer area from the captured image and then detecting the trailer ID area from the trailer area, it is possible to prevent, for example, a character string written on a tractor from being mistakenly detected as a trailer ID, thereby improving the detection accuracy of the trailer ID.

[0077] Although the embodiments have been described above with reference to the accompanying drawings, the present disclosure is not limited to such examples. It is clear that a person skilled in the art can conceive of various modifications, alterations, substitutions, additions, deletions, and equivalents within the scope of the claims, and it is understood that these also fall within the technical scope of the present disclosure. Furthermore, the components in the above-described embodiments may be combined in any manner without departing from the spirit of the invention. [Industrial Applicability]

[0078] The technology of the present disclosure is useful for detecting the trailer ID written on the trailer. [Explanation of symbols]

[0079] 1 yard 2 Entrance area 3 Exit Area 4 Parking Area 10 vehicles 11 Tractor 12 Trailer 20 Management System 21 Imaging device 22 Management device 23 Communication Network 30 Detection device 31 Vehicle Recognition Unit 32 Trailer ID area detection unit 33 Character recognition section 34 Resize section 41 Trailer area detection unit 42 Tractor area detection unit 100 captured images 101, 101A, 101B, 201 Trailer area 102 Tractor area 110 Trailer ID Area 200 resized images 300 graphs 1001 processor 1002 memory 1003 Storage 1004 Input Device 1005 Display device 1006 Communication equipment

Claims

1. a trailer area detection unit that detects a trailer area, which is an area in which the trailer is photographed, from a captured image of at least a portion of a vehicle having a tractor and a trailer coupled thereto; an ID area detection unit that detects an ID area from the trailer area, which is an area where a trailer ID or a container ID is presumed to be written, and outputs information indicating the ID area to a character recognition unit that recognizes the trailer ID or the container ID by character recognition of the ID area; Detection device.

2. the trailer area detection unit detects an area in which the front or rear of the trailer is photographed from the captured image as the trailer area; The detection device according to claim 1 .

3. a resizing unit that resizes the captured image based on the size of the trailer area to generate a resized image, the ID area detection unit detects the ID area from the trailer area of ​​the resized image.

3. The detection device according to claim 1 or 2.

4. the resizing unit resizes the captured image based on the height of the trailer area. The detection device according to claim 3 .

5. The resizing unit If the height of the trailer area is equal to or greater than a predetermined threshold, the captured image is reduced in size; If the height of the trailer area is less than the threshold, the captured image is not reduced. The detection device according to claim 4.

6. the ID area detection unit excludes areas other than the trailer area from detection targets for the ID area. The detection device according to claim 1 .

7. Detecting a trailer area, which is an area in which the trailer is photographed, from a photographed image of at least a part of a vehicle having a tractor and a trailer coupled thereto; Detecting an ID area from the trailer area, which is an area where a trailer ID or a container ID is presumed to be written; outputting information indicating the ID area to a character recognition unit that recognizes the trailer ID or the container ID by character recognition of the ID area; Detection method.

Images