GATT ship launch management system and launch management method
A camera-based system with machine learning determines the open/closed state of grab buckets on GATT ships, addressing manual errors and sensor installation challenges, enabling efficient loading history recording across varying ship specifications.
Patent Information
- Application Number
- JP2022024572
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2022-02-21
- Publication Date
- 2024-11-13
- Estimated Expiration
- 2042-02-21
AI Technical Summary
Existing loading management systems for GATT ships require manual visual checks and are prone to human error, and installing sensors in grab buckets necessitates time-consuming adjustments based on varying crane specifications, leading to potential failures and malfunctions.
A camera-based system that uses machine learning to determine the open/closed state of the grab bucket and records loading history, utilizing a camera device, position information acquisition, and a computing device to create a predictive model for determining object throws into the construction area, eliminating the need for sensor installation and adjustments.
This system allows for accurate and efficient recording of loading history without sensor installation, reducing labor and time for setup, minimizing errors, and ensuring quick application across different GATT ship specifications.
Smart Images

Figure 0007582987000001 
Figure 0007582987000002 
Figure 0007582987000003
Abstract
Description
[Technical field]
[0001] The present invention relates to a system and method for managing the dumping of objects on a gut ship, and more specifically, to a system and method for managing the dumping of objects on a gut ship that can be easily applied regardless of the specifications of the gut ship used when managing the dumping history of objects dumped using the grab bucket of a crane installed on the gut ship. [Background technology]
[0002] Construction work is being carried out in which rubble and other materials loaded on a gut ship are dumped into the construction area using the grab bucket of the crane mounted on the gut ship. Conventionally, a manager visually checked the location and number of times the crane's grab bucket dumped the materials and recorded the dumping history. This required the appointment of a manager to monitor the dumping work using the grab bucket, and it took a lot of effort to record the dumping history. In addition, there was a concern that visual confirmation by the manager would result in human error, such as oversight.
[0003] Therefore, a dumping management system that mechanically records the dumping history by the crane of the gutter ship has been proposed (see, for example, Patent Document 1). In the dumping management system described in Patent Document 1, a grab opening / closing sensor that detects the actual opening and closing of the grab bucket and the feeding of the grab opening / closing wire is installed in the grab bucket in order to determine whether or not the fill sand has been dumped by the grab bucket. However, construction companies that perform dumping work for riprap and the like often use gutter ships owned by rental companies rather than their own companies, and there are cases where they cannot know in advance the specifications (opening / closing mechanism, structure, etc.) of the grab bucket of the crane equipped on the gutter ship used in the work. Therefore, in the method of installing a grab opening / closing sensor in the grab bucket, it is necessary to adjust the position where the sensor is fixed and the sensitivity of the sensor according to the specifications of the grab bucket of the gutter ship used each time, and the sensor setting work requires a lot of effort and time. In addition, the grab bucket vibrates violently when the grab bucket is landed on the object to be dumped loaded in the hold or when the grab bucket is opened and closed. Therefore, there is also a problem that the sensor installed in the grab bucket is prone to failure and malfunction. [Prior art documents] [Patent documents]
[0004] [Patent Document 1] JP 2013-040472 A Summary of the Invention [Problem to be solved by the invention]
[0005] The object of the present invention is to provide a gutter ship dumping management system and dumping management method that can be easily applied when managing the dumping history of objects dumped using the grab bucket of a crane installed on the gutter ship, regardless of the specifications of the gutter ship used. [Means for solving the problem]
[0006] In order to achieve the above object, the gat boat dumping management system of the present invention is a gat boat dumping management system that manages the dumping history of the work of dumping objects into a construction water area using a grab bucket of a crane mounted on the gat boat, and is equipped with a camera device that sequentially acquires image data on the surface of the construction water area, a position information acquisition means that sequentially acquires planar position information of the grab bucket of the crane, a calculation device to which the image data on the surface of the water and the planar position information are sequentially input, and a monitor connected to the calculation device, and the calculation device uses image data of the grab bucket in a closed state and image data of the grab bucket in an open state as teacher data. the grab bucket of the crane at the time the object was dumped into the work water area when the determination unit determines that the object was dumped into the work water area by the grab bucket of the crane; a recording unit that records the planar position information of the grab bucket of the crane at the time the object was dumped as dumping history information when the determination unit determines that the object was dumped into the work water area by the grab bucket of the crane; and a display unit that displays the dumping history information recorded in the recording unit on the monitor.
[0007] The method for managing the dumping of a gat ship of the present invention is a method for managing the dumping history of the work of dumping objects into a construction water area using a grab bucket of a crane mounted on the gat ship, which includes sequentially acquiring image data on the surface of the construction water area by a camera device, sequentially acquiring planar position information of the grab bucket of the crane by a position information acquiring means, sequentially inputting the image data on the surface of the water and the planar position information into a calculation device, and storing image data of the grab bucket in a closed state and image data of the grab bucket in an open state in the calculation device as teacher data for a mechanical learning. In this case, a prediction model for discriminating the open / closed state of the grab bucket that has been learned from the learning is stored in advance, and the arithmetic device is used to sequentially determine whether or not the object to be thrown into the work waters has been thrown into the work waters by the grab bucket of the crane based on the prediction model and the image data on the water, and when it is determined that the object to be thrown into the work waters has been thrown into the work waters by the grab bucket of the crane, the planar position information of the grab bucket of the crane at the time the object to be thrown into the work waters is recorded as throwing history information, and the recorded throwing history information is displayed on a monitor. Effect of the Invention
[0008] According to the present invention, a prediction model for determining whether the grab bucket is open or closed is machine-learned using image data of the grab bucket in the closed and open states as teacher data, and image data of the surface of the work area sequentially acquired by the camera device, and a calculation device sequentially determines whether the object to be thrown into the work area is thrown into the work area by the grab bucket of the crane. The installation position of the camera device is not particularly limited as long as it is a position where image data of the surface of the work area can be acquired. Therefore, even if a sensor is not installed in the grab bucket of the crane, it is possible to record planar position information at the time when the object to be thrown into the work area is thrown into the work area by the grab bucket as throwing history information and display the throwing history information on a monitor. This makes it possible to easily apply the system to managing the throwing history of the object to be thrown into the work area by the grab bucket of the crane mounted on the gutter ship, regardless of the specifications of the gutter ship (grab bucket) used. [Brief description of the drawings]
[0009] [Figure 1] 1 is an explanatory diagram illustrating a side view of a gutting boat equipped with the gutting boat launch management system of the present invention. FIG. [Diagram 2] FIG. 2 is an explanatory diagram illustrating a schematic configuration of the input management system of FIG. [Diagram 3] 2 is an explanatory diagram illustrating an example of an image displayed on the monitor of FIG. 1. [Figure 4] 1. FIG. 4 is an explanatory diagram illustrating another image displayed on the monitor of FIG. [Diagram 5] 1. FIG. 4 is an explanatory diagram illustrating still another image displayed on the monitor of FIG. [Figure 6] FIG. 10 is an explanatory diagram illustrating the configuration of another embodiment of the gutter launch management system of the present invention. [Figure 7] 7 is an explanatory diagram illustrating an example of an image displayed on the monitor of FIG. 6. [Figure 8] 7 is an explanatory diagram illustrating another image displayed on the monitor of FIG. 6. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will now be described with reference to the embodiments shown in the drawings.
[0011] The dumping management system 1 of the present invention shown in Fig. 1 manages the dumping history of dumping objects 30 (such as riprap and filler sand) into a construction water area using a grab bucket 25 of a crane 22 mounted on a gutting barge 20. The gutting barge 20 has a hold 21 for loading the objects 30 to be dumped, and a crane 22 equipped with a grab bucket 25 is mounted on the deck.
[0012] 1 and 2, the input management system 1 includes a camera device 2, a position information acquisition means 3, a computing device 4, and monitors 5a and 5b. The input management system 1 of this embodiment further includes a repeater 6 and an input means 7 for the computing device 4.
[0013] For example, a digital camera or the like is used as the camera device 2. As illustrated in Fig. 3 and Fig. 4, the camera device 2 sequentially acquires image data WD on the water of the construction water area where the work of dumping the dump object 30 by the grab bucket 25 of the crane 22 is performed. In this embodiment, the camera device 2 is installed on the bridge 26 of the gut barge 20, and the shooting direction of the camera device 2 is set to a direction to shoot the construction water area on the side of the hull of the gut barge 20. Fig. 3 illustrates the image data WD on the water when the grab bucket 25 is closed, and Fig. 4 illustrates the image data WD on the water when the grab bucket 25 is open and the dump object 30 is dumped into the construction water area. The image data WD on the water acquired by the camera device 2 is sequentially input to the calculation device 4.
[0014] The installation positions of the camera devices 2 are not particularly limited as long as they are positions where image data WD on the water of the construction water area can be acquired. The camera devices 2 are not limited to being installed on the gut boat 20, and can also be installed in other locations other than the gut boat 20 (for example, other ships, breakwaters, quays, water structures, etc.). The number of camera devices 2 to be installed and their installation positions can be appropriately determined depending on the structure of the gut boat 20 and the angle of view of the camera devices 2. In cases where blind spots that cannot be captured by a single camera device 2 occur due to the structure of the gut boat 20 or cargo carried on board, the camera devices 2 can be installed in multiple locations to acquire image data WD on the water of the construction water area from multiple directions.
[0015] At night or during bad weather when the construction water area is dark, lighting equipment can be used to take pictures with the camera device 2. For example, an infrared night vision camera or a low-illumination camera (high-sensitivity camera) capable of taking pictures even in dark places can be used as the camera device 2. The low-illumination camera (high-sensitivity camera) is equipped with an image processing function such as a wide dynamic range function (WDR function) that allows taking pictures even in dark places, and can take pictures even in low-illumination construction water areas with an illuminance of about 0.005 Lux.
[0016] The position information acquisition means 3 sequentially acquires planar position information (two-dimensional position coordinate information in a planar view) of the grab bucket 25 of the crane 22. An example of the position information acquisition means 3 is a GNSS receiver that receives planar position information from a global positioning system. In this embodiment, a GNSS receiver is installed as the position information acquisition means 3 at the tip 23a of the boom 23 of the crane 22. Since the grab bucket 25 is suspended below the tip 23a of the boom 23 by the hoisting wire 24, the planar position information of the grab bucket 25 can be acquired by acquiring the planar position information of the tip 23a of the boom 23.
[0017] In this embodiment, the planar position information received by the position information acquisition means 3 (GNSS receiving device) is wirelessly transmitted in sequence to the repeater 6 installed on the bridge 26, and the planar position information received by the repeater 6 is wirelessly transmitted in sequence to the computing device 4 installed in the steering room of the bridge 26. The location where the repeater 6 is installed is not particularly limited, and it can be installed in other locations. The repeater 6 can be installed arbitrarily, and for example, it can be configured so that the planar position information acquired by the position information acquisition means 3 (GNSS receiving device) is directly input to the computing device 4 without installing the repeater 6. The position information acquisition means 3 is not limited to this embodiment and can have various other configurations as long as it is configured to acquire the planar position information of the grab bucket 25.
[0018] A computer or the like is used for the calculation device 4. In this embodiment, a monitor 5a for the manager is placed inside the bridge 26, and a monitor 5b for the operator is placed in the driver's cab 22a of the crane 22. The operator's monitor 5b can be provided as desired. Examples of the input means 7 include a mouse, a keyboard, and a touch panel function provided on the monitor 5a. The calculation device 4, the monitor 5a, and the input means 7 can also be configured, for example, with a notebook personal computer or a multi-function mobile terminal (tablet, etc.).
[0019] 2, the camera device 2, monitors 5a, 5b, and input means 7 are communicatively connected to the computing device 4. The position information acquisition means 3 is communicatively connected to the computing device 4 via a repeater 6. The computing device 4 sequentially receives image data WD on the water surface acquired by the camera device 2 and planar position information of the grab bucket 25 acquired by the position information acquisition means 3. The installation position of the computing device 4 is not particularly limited.
[0020] The calculation device 4 has a storage unit 4a, a determination unit 4b, a recording unit 4c, and a display unit 4d as functional elements. The calculation device 4 of this embodiment further has a target range setting unit 4e as a functional element. The storage unit 4a stores a prediction model (computer program) for determining the open / closed state of the grab bucket 25, which is machine-learned using image data of the grab bucket 25 in a closed state and image data of the grab bucket 25 in an open state as teacher data.
[0021] The storage unit 4a stores a large number of image data obtained by photographing the grab bucket 25 in a closed state and the grab bucket 25 in an open state from various angles as teaching data. The image data of the grab bucket 25 stored in the storage unit 4a is not limited to image data obtained by photographing the grab bucket 25 of the crane 22 that manages the loading history, and may be image data of another grab bucket 25.
[0022] A construction company that uses the gut ship 20 to dump rubble and the like often uses gut ships 20 owned by rental companies rather than by its own company. Therefore, there are cases where the specifications of the gut ship 20 and the specifications (type, size, etc.) of the grab bucket 25 equipped on the crane 22 of the gut ship 20 cannot be known in advance. The grab buckets 25 used in dumping work can be mainly classified into orange peel type having many claws and clamshell type having a pair of shells on the left and right, but the shapes of the claws and shells are diverse. Therefore, when the specifications of the grab bucket 25 to be used are not known in advance, it is preferable to store image data of grab buckets 25 of various specifications in the memory unit 4a in advance. The image data of the grab bucket 25 can also be categorized by type, such as orange peel type and clamshell type, and stored.
[0023] The orientation of the claws and shells of the grab bucket 25, the relative positional relationship between the claws and the shells, etc., are different between the closed state and the open state of the grab bucket 25. Therefore, by using image data of the grab bucket 25 in the closed state and image data of the grab bucket 25 in the open state as teacher data, it is possible to generate a prediction model for determining the open / closed state of the grab bucket 25 by machine learning the difference between the closed state and the open state of the grab bucket 25 with an artificial intelligence (AI). Examples of the machine learning method by the storage unit 4a include deep learning, neural network, regression, clustering, and pattern matching. For example, a separate prediction model can be generated and stored in the storage unit 4a for each type of grab bucket 25, such as an orange peel type or a clamshell type.
[0024] The determination unit 4b sequentially determines whether or not the object to be thrown 30 has been thrown into the work water area by the grab bucket 25, based on the above-mentioned prediction model stored in advance in the storage unit 4a and the image data WD of the water above the work water area input from the camera device 2. More specifically, when the determination unit 4b detects, based on the prediction model, that the grab bucket 25 located above the water in the work water area has changed from a closed state to an open state, the determination unit 4b determines that the object to be thrown 30 has been thrown into the work water area by the grab bucket 25. When the grab bucket 25 has changed from an open state to a closed state, the determination unit 4b does not determine that the object to be thrown 30 has been thrown into the work water area by the grab bucket 25.
[0025] When the determination unit 4b determines that the object 30 to be thrown into the construction water area is thrown by the grab bucket 25, the recording unit 4c records the planar position information of the grab bucket 25 at the time when the object 30 to be thrown into the construction water area is thrown as throwing history information. More specifically, as illustrated in FIG. 5, the recording unit 4c divides the construction water area into a plurality of sections S set in advance, and records the number of times the object 30 to be thrown into each section S by the grab bucket 25 as throwing history information. In this embodiment, the construction water area is divided into a plurality of sections S in a lattice shape in a plan view, and the number of times each section S is thrown into is recorded as throwing history information. The method of dividing the construction water area and the number of sections S can be set in advance to desired conditions by the input means 7 according to the size of the construction water area, the required construction accuracy, and the like.
[0026] The display unit 4d displays the input history information recorded in the recording unit 4c on the monitors 5a and 5b, as shown in Fig. 5. In this embodiment, the monitors 5a and 5b are configured to display planar view data PD showing the number of inputs for each section S in a plan view as the input history information. The monitors 5a and 5b are further configured to display image data WD on the surface of the construction water area acquired by the camera device 2 in real time.
[0027] The target range setting unit 4e sets a target range J for judgment by the judgment unit 4b in the above-water image data WD of the construction water area acquired by the camera device 2. In this embodiment, the target range J for judgment by the judgment unit 4b can be set to any range in the above-water image data WD acquired by the camera device 2. As illustrated in Figs. 3 and 4, in this embodiment, the target range J for judgment by the judgment unit 4b can be set on the above-water image data WD displayed on the monitor 5a. In Figs. 3 and 4, the outer frame showing the target range J for judgment is indicated by a dashed line.
[0028] Next, a method for controlling the launch of a gut ship 20 using this launch control system 1 will be described.
[0029] In setting up the dumping management system 1 on the gut boat 20, the camera device 2 is installed in a position where it can capture images of the water surface of the construction water area into which the dumping object 30 is to be dumped. The camera device 2 may be installed on the gut boat 20, or may be installed in a different location other than the gut boat 20 (e.g., another ship, a breakwater, a quay, an above-water structure, etc.). Furthermore, the position information acquisition means 3 (GNSS receiver) is installed at the tip 23a of the boom 23 of the crane 22, and the repeater 6 is installed at a predetermined position on the gut boat 20. The calculation device 4, the monitor 5a, and the input means 7 are arranged inside the bridge 26 where the manager manages the dumping work. In addition, the monitor 5b is arranged in the driver's cab 22a of the crane 22.
[0030] The manager uses the input means 7 to set the division of the construction water area into sections S and the type of grab bucket 25 to be judged in the calculation device 4 (memory unit 4a, judgment unit 4b, recording unit 4c). The method of dividing the construction water area, the number of sections S, the size and shape of each section S, etc. can be appropriately set on the screen of the monitor 5a using the input means 7.
[0031] In the setting screen for the type of grab bucket 25, for example, options such as orange peel type and clamshell type indicating the type of grab bucket 25 are displayed on the monitor 5a, and the administrator uses the input means 7 to select and set the type of grab bucket 25 to be determined from the options. The determination unit 4b selects a prediction model corresponding to the set type of grab bucket 25 from among the pre-stored prediction models and uses it for determination. The type of grab bucket 25 to be determined can also be configured such that the determination unit 4b mechanically identifies the type of grab bucket 25 shown in the above-water image data WD based on the above-water image data WD of the construction water area input from the camera device 2 and image data of various grab buckets 25 pre-stored as teacher data, and the determination unit 4b automatically selects and uses a prediction model corresponding to the identified type of grab bucket 25.
[0032] It is also possible to generate a prediction model by machine learning without distinguishing between types of grab buckets 25, and configure the determination unit 4b to always use the same prediction model without selecting the type of grab bucket 25 to be determined.
[0033] The manager further uses the input means 7 to set the calculation device 4 (target range setting unit 4e) the range J to be judged by the judgment unit 4b in the above-water image data WD. The judgment target range J is set to the entire range or a part of the construction water area into which the throw-in object 30 is thrown. Areas into which the throw-in object 30 is not thrown, specifically, for example, the surface of the gutter boat 20, water areas outside the construction water area, land, etc., are not included in the judgment target range J and are excluded from the judgment.
[0034] As illustrated in Figures 3 and 4, in this embodiment, the above-water image data WD of the construction water area acquired by the camera device 2 is displayed on the monitor 5a, and the judgment target range J by the judgment unit 4b can be set on the above-water image data WD displayed on the monitor 5a. More specifically, the judgment target range J can be set to a desired range on the above-water image data WD by using the input means 7 to change the position and size of the outer frame indicating the judgment target range J displayed superimposed on the above-water image data WD. This completes the setup work of the input management system 1.
[0035] When starting the work of dumping the object 30 to be dumped by the grab bucket 25 of the crane 22 and managing the dumping history, the camera device 2 sequentially acquires image data WD of the above water in the construction water area, and the position information acquisition means 3 sequentially acquires planar position information of the grab bucket 25 of the crane 22. Then, the acquired above-water image data WD and planar position information are each sequentially input to the calculation device 4.
[0036] Then, using the arithmetic unit 4, it is successively determined whether or not the object to be thrown 30 has been thrown into the work water area by the grab bucket 25 of the crane 22, based on a prediction model for determining the open / closed state of the grab bucket 25, which is stored in advance in the memory unit 4a, and on the basis of the image data WD of the water surface input from the camera device 2. Then, when the determination unit 4b determines that the object to be thrown 30 has been thrown into the work water area by the grab bucket 25 of the crane 22, the recording unit 4c records the planar position information of the grab bucket 25 of the crane 22 at the time when the object to be thrown 30 was thrown as throwing history information. Then, the display unit 4d causes the recorded throwing history information to be displayed on the monitors 5a, 5b.
[0037] As illustrated in Fig. 3 and Fig. 4, the monitors 5a and 5b sequentially display the image data WD of the surface of the construction water area acquired by the camera device 2. Also, as illustrated in Fig. 5, the monitors 5a and 5b display the input history information recorded in the recording unit 4c. In this embodiment, the monitors 5a and 5b display plan view data PD (grid diagram) showing the construction water area in a plan view as the input history information, and the plan view data PD displays the boundary line L dividing the section S and information indicating the number of inputs for each section S. Fig. 5 illustrates a case where numbers are displayed as information indicating the number of inputs for each section S, but the information indicating the number of inputs for each section S can also be configured to be distinguishable by, for example, a different color or pattern for each input.
[0038] The input history information shown in Fig. 5 can be displayed superimposed on the above-water image data WD of the construction water area, or can be displayed in a different position without superimposing on the above-water image data WD. The manager manages the construction status and input history of the input work by checking the above-water image data WD of the construction water area displayed on the monitor 5a.
[0039] In this embodiment, the display unit 4d is further configured to sequentially display the determination result of the open / closed state of the grab bucket 25 by the determination unit 4b superimposed on the above-water image data WD displayed on the monitor 5a (5b). Specifically, a frame indicating the range detected as the grab bucket 25 by the determination unit 4b in the above-water image data WD of the construction water area and an open / closed identification display ID indicating the identification result of whether the grab bucket 25 is in a closed state (CLOSE) or an open state (OPEN) are superimposed on the above-water image data WD. The manager can check whether the determination by the determination unit 4b is being performed normally by checking the open / closed identification display ID displayed on the monitor 5a.
[0040] In this way, in the throwing management system 1, by using a prediction model for determining the open / closed state of the grab bucket 25 from the image data of the grab bucket 25 generated in advance by machine learning and the image data WD of the water surface of the construction water area acquired by the camera device 2, it becomes possible to accurately determine whether the throwing object 30 has been thrown into the construction water area by the grab bucket 25 without installing a sensor in the grab bucket 25 of the crane 22. Then, by recording the planar position information at the time when the throwing object 30 is thrown by the grab bucket 25 as throwing history information and displaying the throwing history information on the monitor 5a, the manager can easily manage the throwing history of the throwing object 30 by the grab bucket 25 of the crane 22 of the gutter ship 20. Since the throwing management system 1 can mechanically record the throwing history of the throwing object 30 by the grab bucket 25, the manager's labor can be significantly reduced compared to when the manager records by visual confirmation.
[0041] Furthermore, although construction companies that carry out the loading work often use gutters owned by rental companies rather than their own companies, this loading management system 1 can be easily applied regardless of the specifications of the gutter ship 20 (grab bucket 25). That is, in the conventional method of installing a grab opening / closing sensor in the grab bucket 25, it is necessary to adjust the position where the sensor is fixed and the sensitivity of the sensor each time according to the specifications of the gutter ship 20 (grab bucket 25), but this loading management system 1 does not require such time-consuming and labor-intensive setting work. Therefore, the setting work before construction can be completed more easily and in a short time, and the labor required for the setting work can be significantly reduced. Since the loading work by the crane 22 can be started more quickly, it is also advantageous to shorten the construction period of the loading work.
[0042] Furthermore, in this input management system 1, there is no need to install a sensor on the grab bucket 25, and each device constituting the input management system 1 (such as the camera device 2) can be easily installed in a position with less vibration. Therefore, there is a low risk of failure or malfunction of each device constituting the input management system 1. Therefore, this input management system 1 is very useful for construction companies.
[0043] As in this embodiment, the recording unit 4c divides the construction water area into a number of pre-set sections S and records the number of times the grab bucket 25 of the crane 22 has dropped the object 30 into each section S as drop history information, and the display unit 4d displays information indicating the number of times the object 30 has been dropped into each section S on the monitor 5a (5b), making it very easy to grasp the position where the object 30 was dropped and the number of times it was dropped.
[0044] If the monitor 5a is configured to display the image data WD of the surface of the construction water area acquired by the camera device 2 together with the input history information recorded by the recording unit 4c, the manager can grasp the input history information of the input work as well as the construction status of the input work in real time just by looking at the monitor 5a. Therefore, it is more advantageous to reduce the burden on the manager required for managing the input work.
[0045] If the monitor 5b placed in the driver's cab 22a of the crane 22 is configured to display the input history information recorded by the recording unit 4c, in the same way as the monitor 5a for the manager, the operator of the crane 22 can share and understand the input history information together with the manager, making it easier to carry out input work more smoothly.
[0046] If the determination unit 4b is configured to be able to set the range J to be determined in any range of the on-water image data DW, even if the on-water image data DW acquired by the camera device 2 includes the top of the gutter ship 20, which is not a construction water area, the determination unit 4b can limit the range J to the construction water area and set other ranges that do not correspond to the construction water area as non-determination targets that are not determined by the determination unit 4b. This can reduce the risk of erroneously recording the opening and closing operation of the grab bucket 25 when the grab bucket 25 grabs the object to be thrown 30 loaded in the hold 21 as the work of throwing the object to be thrown 30 into the construction water area. In addition, it is possible to easily and quickly set the range to be recorded as the throwing history by simply setting the determination range J without finely adjusting the shooting direction and angle of view when setting up the camera device 2. This is therefore more advantageous in reducing the time and effort required for setting up the throwing management system 1.
[0047] Furthermore, as in this embodiment, by configuring the determination target range J to be set on the above-water image data DW displayed on the monitor 5a, it becomes possible to set the determination target range J more intuitively on the above-water image data DW, making it easier to set the determination target range J. Also, by directly setting the determination target range J on the above-water image data DW, it is possible to relatively simply configure a computer program that identifies whether or not the grab bucket 25 is located within the determination target range J in the above-water image data DW.
[0048] The method of setting the judgment target range J is not limited to the above-mentioned method, and for example, the judgment target range J can be set on a plane coordinate diagram illustrating the gutter boat 20 and its surrounding water area in a planar view. In this case, for example, the judgment unit 4b judges the range on the water image data WD corresponding to the judgment target range J set on the plane coordinate diagram as the judgment target range J based on the installation position and shooting direction of the camera device 2. Also, for example, the judgment unit 4b is configured to make a judgment only when the planar position information of the grab bucket 25 acquired by the position information acquisition means 3 is on the coordinates within the judgment target range J set on the plane coordinate diagram.
[0049] The input management system 1 can be configured, for example, as shown in Fig. 6. In this embodiment, in addition to the input management system 1 of the embodiment exemplified above, it further includes an AR display camera device 8 and a compass 9. The other configurations are the same as those of the input management system 1 of the embodiment exemplified above.
[0050] The camera device 8 for AR display is installed near the camera device 2 with the same shooting direction as the camera device 2, and sequentially acquires image data of the surface of the water in the construction water area. The compass 9 is installed on the gutt boat 20, and sequentially acquires planar position information and bow direction information of the gutt boat 20. The camera device 8 for AR display and the compass 9 are each communicatively connected to the calculation device 4 (display unit 4d). The image data of the surface of the water in the construction water area acquired by the camera device 8 for AR display and the planar position information and bow direction information of the gutt boat 20 acquired by the compass 9 are sequentially input to the calculation device 4.
[0051] 7 and 8, the display unit 4d of the computing device 4 is configured to display on the monitor 5a the boundary lines L dividing each section S and information indicating the number of times of casting for each section S, superimposed on the above-water image data DW, based on the above-water image data of the construction waters input from the AR display camera device 8 and the position information and bow direction information of the gutter boat 20 input from the compass 9. In this embodiment, the number of times of casting for each section S is made identifiable by a pattern (hatching), but the number of times of casting can also be made identifiable by different colors or shades of color applied to the sections S, numbers, etc.
[0052] By configuring the monitors 5a, 5b to display such augmented reality images, the input history information can be more intuitively grasped on images showing the actual construction situation, making it easier to manage the input work and give work instructions to operators, which is even more advantageous in improving the construction efficiency of the input work.
[0053] In this embodiment, the predicted drop position display FP is further configured to be displayed on the water surface position where the thrown object 30 will fall when the grab bucket 25 is opened, superimposed on the above-water image data WD of the construction water area. In Figs. 7 and 8, a projected mark indicating the center of the grab bucket 25 is displayed on the water surface as the predicted drop position display FP, but for example, the predicted drop position display FP can be configured to display (for example, a color or pattern) the section S in which the grab bucket 25 is located differently from other sections S. In this way, when the predicted drop position display FP is displayed superimposed on the above-water image data WD, it becomes easier to grasp in which section S the grab bucket 25 is located, so that the manager can more easily give work instructions to the operator of the crane 22, and the operator can more easily align the grab bucket 25.
[0054] In the embodiment exemplified above, the determination unit 4b performs a determination on the above-water image data WD acquired by one camera device 2, but for example, the determination unit 4b may perform a determination on each of the above-water image data WD acquired by two or more camera devices 2. In this case, when the determination unit 4b determines that the throw-in object 30 has been thrown into the construction water area by the grab bucket 25 of the crane 22 in at least one of the above-water image data WD acquired by the multiple camera devices 2, the recording unit 4c records the planar position information of the grab bucket 25 of the crane 22 at the time when the throw-in object 30 was thrown in as throw-in history information. In this way, even when there is a blind area that cannot be photographed by one camera device 2, it is possible to more reliably record the throw-in history by photographing the blind area with another camera device 2.
[0055] In the embodiment exemplified above, the judgment target range J for judgment by the judgment unit 4b is set, but for example, a non-judgment target range for which judgment by the judgment unit 4b is not performed is set, so that the range for judgment by the judgment unit 4b in the image data WD on the water is limited. In the embodiment exemplified above, the calculation device 4 and the manager's monitor 5a are placed on the bridge 26 of the gut boat 20, but for example, the calculation device 4 and the manager's monitor 5a can be placed outside the gut boat 20. In this case, for example, the image data WD on the water acquired by the camera device 2 and the planar position information of the grab bucket 25 acquired by the position information acquisition means 3 are sequentially transmitted to the calculation device 4 placed outside the gut boat 20 using an Internet line or the like, so that the throwing work can be managed from outside the gut boat 20. [Explanation of symbols]
[0056] 1 Input Management System 2 Camera equipment 3 Location information acquisition means 4 Arithmetic unit 4a Storage section 4b Judgment part 4c Recording section 4d display 4e Target range setting section 5a, 5b Monitor 6 Repeater 7. Input Methods 8. AR display camera device 9 Direction meter 20 Gutt Ship 21 Hold 22 Crane 22a Driver's cab 23. Boom 23a Tip 24 Suspension wire 25 Grab Bucket 26 Funabashi 30 Input items WD Water image data J. Judgment scope ID Open / Close Identification Display FP predicted fall position display PD Planar Data L border S Section
Claims
1. In a gat barge dumping management system that manages the dumping history of dumping objects into a construction water area using a grab bucket of a crane mounted on the gat barge, The system includes a camera device for sequentially acquiring image data on the surface of the work area, a position information acquiring means for sequentially acquiring planar position information of the grab bucket of the crane, a computing device to which the image data on the surface of the water and the planar position information are sequentially input, and a monitor connected to the computing device, The arithmetic device has a memory unit in which a prediction model for determining the open / closed state of the grab bucket, which has been machine-learned using image data of the grab bucket in a closed state and image data of the grab bucket in an open state as teacher data, a judgment unit for sequentially judging whether the object to be thrown into the work water area has been thrown by the grab bucket of the crane based on the prediction model and the image data on the water, a recording unit for recording, as throwing history information, the planar position information of the grab bucket of the crane at the time the object to be thrown into the work water area was thrown when the judgment unit judges that the object to be thrown into the work water area has been thrown by the grab bucket of the crane, and a display unit for displaying the throwing history information recorded in the recording unit on the monitor.
2. 2. A gutter boat throwing management system as described in claim 1, wherein the recording unit divides the construction waters into a plurality of pre-set sections and records the number of times the grab bucket of the crane has thrown the object into each section as the throwing history information, and the display unit displays information indicating the number of times the object has been thrown into each section on the monitor.
3. A gutting boat casting management system as described in claim 2, wherein the display unit displays on the monitor the boundary lines dividing each of the sections and information indicating the number of times each section is cast, superimposed on the image data on the water.
4. A system for managing the launch of a gutting boat according to any one of claims 1 to 3, wherein the range to be judged by said judging unit can be set to any range of the image data on the water.
5. A gutting boat launch management system as described in claim 4, wherein the judgment target range can be set on the image data on the water displayed on the monitor.
6. A method for managing the dumping history of a gutter ship for dumping objects into a construction water area using a grab bucket of a crane mounted on the gutter ship, comprising: The camera device sequentially acquires image data on the surface of the work area, and the position information acquisition means sequentially acquires planar position information of the grab bucket of the crane, and the image data on the surface of the water and the planar position information are sequentially input into a calculation device, The calculation device pre-stores a prediction model for determining the open / closed state of the grab bucket, which has been machine-learned using image data of the grab bucket in a closed state and image data of the grab bucket in an open state as teacher data, and the calculation device is used to sequentially determine whether or not the object to be thrown into the work waters has been thrown by the grab bucket of the crane based on the prediction model and the image data on the water, and when it is determined that the object to be thrown into the work waters has been thrown by the grab bucket of the crane, the calculation device records the planar position information of the grab bucket of the crane at the time the object to be thrown into the work waters as throwing history information, and displays the recorded throwing history information on a monitor.
Citation Information
Patent Citations
Riprap throw-in management system
JP2010261228A
Caisson filler feed management system and method, and computer program therefor
JP2013040472A
Support system of stacking work
JP2014172754A
Height management device of water bottom construction area
JP2017066642A
Control system, information processing unit, and information processing method
JP2021147196A
