Stagnation detection system, stagnation analysis apparatus and program

Abstract

To early detect the stagnation of an object to be fed at each transit part of belt conveyors and to prevent the delay and stop of the conveyance operation of the object to be fed due to the stagnation of the object to be fed even if the number of monitoring operators is small.SOLUTION: A stagnation detection system comprises: plural belt conveyors B1 to B4 for conveying an object to be fed; image capturing means C1 to C3 arranged at each of transit parts T1 to T3 between the plural belt conveyors; first analysis means for analyzing whether the object to be fed is normally flowing or not at the transit parts based on video data captured by the image capturing means; and alert issue means 6 for issuing an alert when it is determined that the stagnation of the object to be fed is generated in each transit part. The first analysis means extracts the border line of the object to be fed from the video data to analyze the flow of the object to be fed, and determines that the object to be fed is stagnated at each transit part when the border line of a vector different from the border line when the object to be fed is normally flowing is extracted.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a belt conveyor congestion detection system, a congestion analysis device, and a program. [Background technology]

[0002] For example, when transporting excavated soil (mud) from a face to an opening at an excavation site such as a tunnel, if the transport distance is long, multiple belt conveyors are connected in series to ensure the driving force. However, when multiple belt conveyors are connected in series, a transfer section is provided between the upstream belt conveyor and the downstream belt conveyor to transfer the soil, but the soil is likely to accumulate in such a transfer section. If the soil accumulates, a blockage may occur, and the soil transport operation may be stopped. Therefore, for example, as disclosed in Patent Document 1, a technology for photographing the soil transported by the belt conveyor is used to monitor the transport status of the soil at the transfer section. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 08-244954 Summary of the Invention [Problem to be solved by the invention]

[0004] Images of the soil and sand captured at multiple transfer points are sent to a central control room, where monitoring workers stationed there monitor the images. However, if the soil and sand are transported over a long distance, the number of transfer points also increases, and if the number of monitoring workers is small, it is not possible to monitor the situation at all transfer points, and detection of soil and sand accumulation at transfer points may be delayed. If soil and sand accumulates at transfer points, soil transportation operations will be halted, so there is a need to develop technology that can detect soil and sand accumulation at transfer points early, even with a small number of monitoring workers.

[0005] The present invention has been made in consideration of the above-mentioned circumstances, and aims to enable early detection of backlogs of supplied materials at transfer sections of a conveyor belt, even with a small number of monitoring workers, and to prevent delays or stoppages in the transportation of the supplied materials due to the backlogs of the supplied materials. [Means for solving the problem]

[0006] In order to solve the above problems, the invention described in claim 1 is a retention detection system, A plurality of belt conveyors for transporting the objects to be supplied to the downstream side; An imaging means is disposed at each transfer section between the plurality of belt conveyors, and captures an image of the object flowing at the transfer section; a first analysis means for acquiring image data captured by the imaging means and analyzing whether the object is flowing normally at the transfer section based on the image data; and an alert issuing means for issuing an alert when it is determined that the object is being supplied at the transfer section as a result of the analysis by the first analysis means, The first analysis means extracts a contour of the object from the video data and analyzes the flow of the object, and when a contour of a vector different from the contour when the object is flowing normally is extracted, it determines that the object is stuck at the transfer area.

[0007] The invention described in claim 2 is the retention detection system described in claim 1, The image data is input, and information regarding the retention of the object at the transfer section is output as a machine-learned model, The first analysis means is characterized in that it uses the trained model to analyze whether the supplied item is flowing normally at the transfer area.

[0008] The invention described in claim 3 is the retention detection system described in claim 1, The system further comprises a second analysis means for analyzing whether the supplied goods are flowing normally at the transfer section based on the acquired video data, and the analysis by the second analysis means is performed by a method different from that of the first analysis means.

[0009] In order to solve the above problems, the present invention provides a retention analysis device, comprising: A photographing means arranged at each transfer section between the plurality of belt conveyors photographs an image of the state in which the supplied object flows at the transfer section, an image acquisition means for acquiring image data captured by the imaging means; an analysis means for analyzing whether the object is flowing normally at the transfer section based on the image data acquired by the image acquisition means; and an alert issuing means for issuing an alert when it is determined that the object is being supplied at the transfer section as a result of the analysis by the analysis means, The analysis means extracts a contour of the object from the video data and analyzes the flow of the object, and when a contour of a vector different from the contour when the object is flowing normally is extracted, it determines that the object is stuck at the transfer area.

[0010] In order to solve the above problems, the invention described in claim 5 is characterized in that a photographing means arranged at each transfer section between a plurality of belt conveyors photographs an image of the state in which the objects to be supplied flow at the transfer section, The computer of the retention analysis device, an image acquisition means for acquiring image data captured by the imaging means; an analysis means for analyzing whether the object is flowing normally at the transfer section based on the image data acquired by the image acquisition means; an alert issuing means for issuing an alert when it is determined that the object is being held up in the transfer section as a result of the analysis by the analysis means; A program that functions as The analysis means extracts a contour of the object from the video data and analyzes the flow of the object, and when a contour of a vector different from the contour when the object is flowing normally is extracted, it determines that the object is stuck at the transfer area. Effect of the Invention

[0011] According to the present invention, even with a small number of monitoring workers, it is possible to detect early on the backlog of supplied materials at transfer sections of the belt conveyor, and to prevent delays or stoppages in the transportation work of the supplied materials due to the backlog of the supplied materials. [Brief description of the drawings]

[0012] [Figure 1] FIG. 1 is a diagram illustrating an overview of a belt conveyor congestion detection system. [Diagram 2] FIG. 2 is a block diagram showing a configuration of a retention analysis device. [Diagram 3] FIG. 11 is a diagram illustrating an example of an analysis result of an acquired image. [Figure 4] FIG. 11 is a diagram illustrating an example of an analysis result of an acquired image. [Diagram 5] FIG. 11 is a diagram illustrating an example of an analysis result of an acquired image. [Figure 6] FIG. 11 is a diagram illustrating an example of an analysis result of an acquired image. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, the embodiments of the present invention will be described with reference to the drawings. However, the embodiments described below are subject to various technically preferable limitations for carrying out the present invention, but the technical scope of the present invention is not limited to the following embodiments and illustrated examples.

[0014] 1 is a diagram for explaining an outline of the belt conveyor retention detection system, in which the reference characters B1 to B4 respectively indicate belt conveyors. The multiple belt conveyors B1 to B4 supply objects from upstream to downstream. In this embodiment, the object to be supplied and transported from upstream to downstream by the belt conveyors B1 to B4 is excavated earth and sand (mud) discharged from an excavation site such as a tunnel. However, the object is not limited to this, and may be, for example, aggregate produced in an aggregate plant or earth and sand collected from an earth and sand mining site.

[0015] In order to minimize the height difference, the multiple belt conveyors B1 to B4 may be arranged diagonally. In other words, by arranging one belt conveyor B2 (B3) diagonally so that the upstream end is low and the downstream end is high, it is possible to install the multiple belt conveyors B1 to B4 even on a level ground. Any of the plurality of belt conveyors B1 to B4 may be a self-propelled belt conveyor.

[0016] When a plurality of belt conveyors B1 to B4 are used, transfer sections T1 to T3 for transferring the supplied articles are provided between the belt conveyors B1 to B3 located on the upstream side and the belt conveyors B2 to B4 located on the downstream side. That is, the upstream ends of the downstream belt conveyors B2 to B4 are disposed below the downstream ends of the upstream belt conveyors B1 to B3, forming areas where the objects to be supplied transfer from the upstream to the downstream belt conveyors B2 to B4. These areas are called transfer areas T1 to T3.

[0017] Hoppers H1-H3 are installed at transfer sections T1-T3 to prevent the soil and sand being fed from scattering. The downstream ends of the belt conveyors B1-B3 located on the upstream side are arranged above the hoppers H1-H3, and the upstream ends of the belt conveyors B2-B4 located downstream are arranged below the hoppers H1-H3. Hoppers H1 to H3 are gradually tapered from top to bottom, making it easy for them to receive objects conveyed by belt conveyors B1 to B3 located upstream and to place the objects on belt conveyors B2 to B4 located downstream. However, the hoppers H1 to H3 do not necessarily have to be provided. If the hoppers H1 to H3 are not provided, it is sufficient to provide side walls on both sides of the belt conveyors B1 to B4, for example, so long as the objects to be supplied can be prevented from scattering.

[0018] At the transfer sections T1 to T3, cameras C1 to C3 are installed which are photographing sections for photographing the state in which the supplied object is being transported. That is, in this embodiment, cameras C1 to C3 are installed above the hoppers H1 to H3 to capture images of the interiors of the hoppers H1 to H3. However, this is not limited to this, and the cameras C1 to C3 may capture images of the items being discharged from the lower ends of the hoppers H1 to H3 from the sides. Furthermore, images may be captured from above and from the sides in parallel.

[0019] The cameras C1 to C3 each include a storage unit that stores image data and a communication unit that transmits the image data stored in the storage unit to an external device. The cameras C1 to C3 are connected to the external device, i.e., the retention analysis device 1, so as to be able to communicate with each other. Therefore, the cameras C1 to C3 can transmit image data showing the transport status of the supplied object to the retention analysis device 1.

[0020] The retention analysis device 1 is a computer that acquires and analyzes image data (video data) from a plurality of cameras C1 to C3 that capture images of transfer sections T1 to T3 on a plurality of belt conveyors B1 to B4. Such a retention analysis device 1 includes a calculation unit 2, a storage unit 3, a display unit 4, a communication unit 5, and an alert issuing unit 6, as shown in FIG.

[0021] The calculation unit 2 is composed of a CPU (Central Processing Unit), a RAM (Random Access Memory), etc. The CPU of the calculation unit 2 reads out various programs stored in the storage unit 3, expands them in the RAM, executes various processes (such as image analysis process described later) according to the expanded programs, and centrally controls the operations of each part of the retention analysis device 1.

[0022] The storage unit 3 is configured with a non-volatile memory, a hard disk, etc. The storage unit 3 stores various programs executed by the calculation unit 2, various data required for executing the programs, etc. The various programs executed by the calculation unit 2 include an image acquisition program, a first analysis program, a second analysis program, and an alert issuance program. These programs function as one means of the retention analysis device 1 and the retention detection system in cooperation with the calculation unit 2, the storage unit 3, the display unit 4, the communication unit 5, and the alert issuance unit 6. Moreover, video data acquired from a plurality of cameras C1 to C3 is stored in the storage unit 3. Note that a database may be prepared separately, and the video data may be stored in the database. Furthermore, the storage unit 3 stores a learned model 3a for analyzing video data, which has been trained in advance using a large amount of video data.

[0023] The display unit 4 is composed of a monitor such as an LCD (Liquid Crystal Display) or a CRT (Cathode Ray Tube), and is configured to display the images captured by the cameras C1 to C3 in real time according to the instructions of the display signal input from the calculation unit 2. The display units 4 may be provided in the same number as the cameras C1 to C3, or the display screen may be divided into multiple parts to display the images captured by the cameras C1 to C3. The display screen displayed on the display unit 4 is viewed and confirmed by a monitoring worker who monitors the accumulation of the supplied objects at the transfer sections T1 to T3. The images displayed on the display unit 4 are moving images captured by the cameras C1 to C3, but it may also be possible to display images undergoing analysis processing by the first analysis program or the second analysis program (see Figs. 3 to 6). In this case, the display unit 4 has a mode switching function, and the monitoring operator can select between a mode in which the analysis processing status is displayed and a mode in which it is not displayed.

[0024] The communication unit 5 is configured by a communication module etc. The communication unit 5 transmits and receives various signals and various data to and from the communication units of the cameras C1 to C3 connected by wire or wirelessly via a communication network. The communication network may include various communication networks such as telephone networks, ISDN networks, optical fibers, mobile communication networks, communication satellite networks, CATV networks, and other dedicated lines, as well as Internet service providers (i.e., the Internet) that connect them. It may also be a collective communication network in which various communication networks such as LANs (Local Area Networks), WANs (Wide Area Networks), WiFi (Wireless Fidelity), Bluetooth (registered trademark), and NFC (Near Field Communication) are connected to each other so that they can communicate with each other. The connection form may be wired, wireless, or a combination of wired and wireless. The communication unit 5 may be communicably connected to a mobile information terminal (not shown). The mobile information terminal is used by a monitoring worker who monitors the accumulation of the supplied object, and is configured, for example, by a smartphone, a tablet terminal, a mobile phone, or the like.

[0025] In this embodiment, the alert issuing unit 6 is an audio output unit constituted by a speaker or the like, and is configured to output an alarm in accordance with the instruction of an alert issuing signal input from the calculation unit 2. However, the type of alert is not limited to this, and may be a light emitting unit constituted by a warning lamp or the like, or may display an alert on the display screen in cooperation with the display unit 4 (for example, a sentence indicating the alert, blinking of the display screen, etc.). In addition, the alert signal from the calculation unit 2 may be transmitted to a portable information terminal used by a monitoring worker via the communication unit 5. In this case, the monitoring worker can check the alert on the portable information terminal.

[0026] (Image acquisition processing) The retention analysis device 1 acquires real-time video data from cameras C1 to C3 arranged at transfer points T1 to T3 between a plurality of belt conveyors B1 to B4. More specifically, the calculation unit 2 in the retention analysis device 1 executes a video acquisition program to receive video data from a plurality of cameras C1-C3 as input via a communication network, and outputs and stores the data in the storage unit 3. At this time, the video data is associated with identification information that identifies at which of the transfer points T1-T3 the cameras C1-C3 are installed. Such an image acquisition program functions as an image acquisition means of the retention analysis device 1 and the retention detection system.

[0027] (First analysis process) The retention analysis device 1 analyzes, based on the stored video data, whether or not the supplied goods are flowing normally at the transfer sections T1 to T3. More specifically, the calculation unit 2 in the retention analysis device 1 executes a first analysis program to analyze the stored video data. In the analysis process using the first analysis program, all of the video data from the multiple cameras C1 to C3 stored in the storage unit 3 is the subject of the analysis process. The analysis results from the first analysis process indicate whether the supplied materials are flowing normally or are stagnating (including clogging) in the transfer sections T1 to T3 (hoppers H1 to H3). Such a first analysis program functions as the first analysis means of the retention analysis device 1 and the retention detection system.

[0028] The analysis process by the first analysis program is performed by a method of extracting the contour line of the object from the video data and analyzing the flow of the object, as shown in FIGS. Since the conveyors B1-B4 transport the objects from upstream to downstream, the first analysis process first analyzes the unidirectional flow of the objects downstream in the video data. When the objects are flowing in one direction, most of the contour lines extend straight downstream, and it is determined that the objects are flowing normally. Note that Figures 3, 4, and 5 are photographs of the flow of supplied materials taken from different angles at different transfer sections T1 to T3, and Figures 3(a), 4(a), and 5(a) each show the analysis results when the supplied materials are flowing normally. In addition, the white arrows in Figures 3(a), 4(a), and 5(a) show the direction from upstream to downstream when the material is flowing normally, and the outline of the material extends straight along these arrows.

[0029] On the other hand, when the material begins to accumulate, it will gradually accumulate while flowing sideways instead of in one direction, eventually resulting in a complete blockage. When the object starts to accumulate, the outline of the object has a vector different from that of the object flowing normally. In other words, an outline unrelated to the outline of the object flowing normally appears. When such an outline starts to appear, it is determined that the object has started to accumulate. Figure 3(b), Figure 4(b), and Figure 5(b) show the analysis results when the supplied material is stagnant. The arc-shaped arrows in Figures 3(b), 4(b), and 5(b) show how the material is transported from above the accumulated material, gradually piling up, and how the material that falls from above slides down the slope of the accumulated material. The contours of the different vectors above extend in the direction of the arc-shaped arrows, and are short because the flow is stagnant. Also, because the material is piling up, it has a slightly mountain-shaped shape.

[0030] Even if the above-mentioned contour lines of different vectors appear when the supply material starts to stagnate, if the contour lines in the case of normal flow are also included, the stagnation in that case is determined to be relatively light. When the above-mentioned contour lines of different vectors appear as the supply material starts to stagnate and the contour lines in the case of normal flow become fewer, the stagnation in that case is judged to be relatively severe. When there is no contour line in the case of normal flow and the contour lines of the different vectors are almost nonexistent, it is determined that the object is clogged. That is, the degree of accumulation of the supplied objects is classified into a normal state where no accumulation occurs, a mild state, a severe state, and a state where a blockage occurs. The degree of accumulation of such supplied objects may be set to an intermediate degree, such as a moderate degree, between mild and severe. The degree of accumulation of such supplied objects is stored in the storage unit 3 as information related to the accumulation of the supplied objects at the transfer sections T1 to T3.

[0031] Moreover, the analysis process by the first analysis program is performed by calculating the retention degree of the supplied object in the video data by an AI (Artificial Intelligence) function. That is, the retention analysis device 1 inputs the video data acquired from the multiple cameras C1 to C3 to the trained model 3a, outputs the retention degree of the supplied object in the video data from the trained model 3a, and can associate the retention degree with the acquired video data. Such a trained model 3a is a neural network trained using teacher data that associates a large number of video data showing various degrees of stagnation with information indicating a large number of states of stagnation that are each associated with these video data, and is stored in the memory unit 3 as described above. The trained model 3a may be trained for each piece of video data acquired from each of the cameras C1 to C3, or may be trained based on all of the video data acquired from each of the cameras C1 to C3.

[0032] In this embodiment, the video data is analyzed using the trained model 3a as described above, but when analyzing the video data without using the trained model 3a, the analysis process is performed based on the number, length, shape, etc. of vector contour lines that differ from the contour lines when the video data is flowing normally.

[0033] (Second analysis process) The stagnation analysis device 1 executes a second analysis program that performs analysis using a method different from that of the first analysis program, and analyzes whether the supplied goods are flowing normally at transfer sections T1 to T3 based on the stored video data. More specifically, the calculation unit 2 in the retention analysis device 1 executes the second analysis program to analyze the stored video data. In the analysis process using the second analysis program, all of the video data from the multiple cameras C1 to C3 stored in the storage unit 3 is the subject of the analysis process. The analysis results from the second analysis process indicate whether the supplied materials are flowing normally or are stagnating (including clogging) in the transfer sections T1 to T3 (hoppers H1 to H3). Such a second analysis program functions as the second analysis means of the retention analysis device 1 and the retention detection system.

[0034] The analysis process by the second analysis program is performed by a method of extracting the movement (displacement) of the object from the video data and analyzing the flow of the object (a so-called optical flow analysis method), as shown in FIG. In the analysis process by the second analysis program, the degree of accumulation of the supplied objects is determined based on the number of vectors (thin lines in Figure 6) representing the movement of the supplied objects, the direction in which they extend, and the degree of concentration.

[0035] In the second analysis process by such a second analysis program, the unidirectional flow of the object toward the downstream in the video data is first analyzed. The white arrow in Fig. 6(a) shows the direction from upstream to downstream when the object is flowing normally. When the object is flowing in one direction, the vector representing the object's movement (the multiple thin lines shown in Fig. 6a) has a large portion extending in the direction of the white arrow, which determines that the object is flowing normally.

[0036] On the other hand, when the material begins to accumulate, it will gradually accumulate while flowing sideways instead of in one direction, eventually resulting in a complete blockage. Figure 6(b) shows the state when the object starts to accumulate, and the vectors (thin lines in Figure 6(b)) that represent the movement of the object at that time are numerous and extend in various directions. Furthermore, the thin lines are concentrated in a specific location, which indicates that the object is accumulating.

[0037] Also, the analysis process by the second analysis program may be performed by calculating the retention degree of the supplied object in the video data by the AI ​​function, as in the case of the analysis process by the first analysis program. That is, the retention analysis device 1 inputs the video data acquired from the multiple cameras C1 to C3 into another trained model different from the trained model 3a used in the analysis process by the first analysis program, outputs the retention degree of the supplied object in the video data from the other trained model, associates the retention degree with the acquired video data, and performs the analysis process using such another trained model.

[0038] The analysis process by the second analysis program and the analysis process by the first analysis program are performed simultaneously, or one of them is performed first. However, in this embodiment, since the analysis accuracy of the first analysis process is higher than that of the second analysis process, the priority of the analysis results is set higher for the first analysis process than for the second analysis process. In addition, it is preferable to prioritize the first analysis process and the second analysis process and reflect the analysis results in the training of the trained model 3a and other trained models, as this can improve the analysis accuracy of these trained models. In particular, since it takes a few seconds for the backlog of supplied material to go from a mild stage to a clogged state, it is preferable to employ not only the first analysis means but also the second analysis means to improve the analysis accuracy using the trained model, in order to improve the speed at which backlog of supplied material is detected.

[0039] (Alert processing) When it is determined in at least the result of the first analysis process out of the first analysis process and the second analysis process that the supplied object is stuck at the transfer sections T1 to T3, the stuck analysis device 1 executes an alert issuance program to issue an alert. To explain in more detail, the calculation unit 2 in the retention analysis device 1 executes an alert issuance program to input an instruction of an alert issuance signal to the alert issuance unit 6, and the alert issuance unit 6 outputs the alert stored in the storage unit 3. The alert in this embodiment is a warning as described above, and the alert issuance unit 6 that has received the instruction outputs the warning. The type of alarm may be changed depending on the degree of retention of the analyzed supply material. Also, when there is a difference in analysis speed between the first analysis process and the second analysis process, the type of alarm issued when the result of the first analysis process is obtained may be different from the type of alarm issued when the result of the second analysis process is obtained. Such an alert issuing program functions as an alert issuing means of the retention analysis device 1 and the retention detection system.

[0040] (Flow of operations by the congestion detection system) When the multiple belt conveyors B1-B4 are operating and transporting the excavated soil as the object to be supplied, the cameras C1-C3 take pictures of the object to be supplied at each of the transfer points T1-T3 (inside the hoppers H1-H3). Based on this premise, the image data taken by the cameras C1-C3 are transmitted one by one to the retention analysis device 1. The retention analysis device 1 stores the video data one by one in the storage unit 3 by the video acquisition means. The video data is then analyzed by the first analysis means and the second analysis means, and as a result, if it is determined that the supplied object is stuck at any of the transfer sections T1 to T3, an alert is issued by the alert issuing means.

[0041] When an alert is issued, the monitoring worker immediately stops the operation of the multiple belt conveyors B1 to B4 and checks the status of the backlog of the supplied objects on the display unit 4. Then, the monitoring worker performs work to eliminate the backlog of the supplied objects at the transfer points T1 to T3 in the relevant locations. In addition, the work of clearing the backlog of supplied materials may be performed manually by a monitoring worker or other worker, or devices for clearing backlogs may be installed at transfer sections T1 to T3 and the backlog of supplied materials may be cleared remotely.

[0042] (Effects of this embodiment) According to this embodiment, the following excellent effects are obtained. That is, the system includes a plurality of belt conveyors B1-B4 that transport the object (excavated earth and sand) to supply the object to the downstream side, cameras C1-C3 that are arranged at each of the transfer sections T1-T3 between the plurality of belt conveyors B1-B4 and capture images of the object flowing at the transfer sections T1-T3, a first analysis means that acquires image data captured by the cameras C1-C3 and analyzes whether the object flows normally at the transfer sections T1-T3 based on the image data, and an alert issuing means 6 that issues an alert when it is determined that the object is stuck at the transfer sections T1-T3 as a result of the analysis by the first analysis means, so that it is possible to analyze whether the object flows normally or is stuck at each of the transfer sections T1-T3 in the plurality of belt conveyors B1-B4. Then, when a stuck object is found at any of the transfer sections T1-T3, the alert issuing means 6 issues an alert to notify the monitoring operator. This makes it possible to detect early on the backlog of supplied objects at the transfer points T1-T3 of the multiple belt conveyors B1-B4 even with a small number of monitoring workers, thereby preventing delays or stoppages in the transportation of the supplied objects due to the backlog of the supplied objects.

[0043] In addition, the first analysis means extracts the contour of the object from the video data to analyze the flow of the object, and when a contour of a vector different from the contour when the object is flowing normally is extracted, it is determined that the object is stuck at the transfer section, so that it is possible to extract the contour of the object that is flowing, the contour of the object that is gradually getting stuck, or the contour of the object that has caused a blockage. This makes it easy to distinguish and analyze whether the object is flowing or not, which contributes to early detection of the stuck object at the transfer sections T1 to T3.

[0044] The system further includes a trained model 3a that has been machine-learned to use video data as input and information regarding the backlog of supplied goods at transfer sections T1 to T3 as output, and the first analysis means uses the trained model 3a to analyze whether the supplied goods are flowing normally at transfer sections T1 to T3.The first analysis means can therefore accurately analyze the video data acquired from the cameras C1 to C3, enabling backlogs of supplied goods to be detected quickly and reliably.

[0045] Also, a second analysis means is provided for analyzing whether the objects are flowing normally at the transfer sections T1-T3 based on the image data acquired from the cameras C1-C3, and the analysis by the second analysis means is performed by a method different from that of the first analysis means, so that both the first analysis means and the second analysis means can perform a doubly analysis of whether the objects are flowing normally at the transfer sections T1-T3. Therefore, the stagnation of the objects at the transfer sections T1-T3 can be accurately detected.

[0046] The program also causes the computer of the stagnation analysis device 1 to function as an image acquisition means for acquiring image data captured by the image acquisition means, an analysis means for analyzing whether the supplied object is flowing normally at the transfer section based on the image data captured by the image acquisition means, and an alert issuing means for issuing an alert when it is determined that the supplied object is stagnating at the transfer section as a result of the analysis by the analysis means. Therefore, the computer of the stagnation analysis device 1 executing such a program can analyze whether the supplied object is flowing normally or is stagnating at each of the transfer sections T1 to T3 of the multiple belt conveyors B1 to B4. When the stagnancy of the supplied object is found at any of the transfer sections T1 to T3, an alert can be issued by the alert issuing means 6 to notify the monitoring worker. This allows the stagnancy of the supplied object at the transfer sections T1 to T3 of the multiple belt conveyors B1 to B4 to be discovered early even with a small number of monitoring workers, and prevents delays and stops in the transport work of the supplied object due to the stagnancy of the supplied object.

[0047] [Modifications] The embodiments to which the present invention can be applied are not limited to the above-described embodiments, and can be modified as appropriate without departing from the spirit of the present invention. Modifications are described below. The following modifications may be combined as far as possible. In each of the following modifications, elements common to the above-described embodiments are denoted by the same reference numerals, and descriptions thereof are omitted or simplified.

[0048] In the above embodiment, the analysis processing by the first analysis program is performed using a method (hereinafter referred to as the first method) of extracting the contour line of the supplied object from the video data and analyzing the flow of the supplied object, and the analysis processing by the second analysis program is performed using a method (hereinafter referred to as the second method) of extracting the movement (displacement) of the supplied object from the video data and analyzing the flow of the supplied object. In this modification, the analysis process by the first analysis program is performed in a manner different from that of the above embodiment (or may be the same as that of the above embodiment). Also, the analysis process by the second analysis program is performed in a manner different from that of the above embodiment and is also performed in a manner different from that of the first analysis program.

[0049] In addition to the above-mentioned first and second methods, the analysis method adopted in this modification includes an analysis method (hereinafter, the third method) that averages the acquired video data over multiple frames. This third method is a method that averages frames over a few seconds, and when the excavated soil is flowing normally as shown in the video data, fine soil particles disappear and become easier to see, while stagnant soil becomes clearly visible.

[0050] Another analysis method is to extract only the foreground from the foreground and background in the acquired video data (hereinafter referred to as the fourth method). This fourth method extracts only the foreground parts of the video data that are changing and moving, making it easy to extract falling soil and sand. On the other hand, it does not extract stagnant soil and sand, so it is possible to focus the analysis on moving soil and sand.

[0051] Another example is an analysis method (hereinafter, the fifth method) that divides parts of the acquired video data that have the same characteristics into multiple regions. This fifth method distinguishes the difference in color between the sediment and the side walls of hoppers H1-H3, and can recognize the boundaries between the sediment and hoppers H1-H3. If the area of ​​the sediment is narrow and remains that way, it can be determined that the sediment is flowing normally, and if the area of ​​the sediment gradually expands, it can be determined that the sediment is stagnating.

[0052] The analysis process by the first analysis program may employ any one of the second to fifth methods in addition to the first method. Furthermore, the analysis process by the second analysis program employs a method among the first to fifth methods that is different from the analysis process by the first analysis program. The combination of the first to fifth methods is considered taking into account various conditions, such as the type of material to be fed and the conveying speed, and the optimal combination of analysis methods is appropriately selected. [Explanation of symbols]

[0053] B1~B4 Conveyor belt T1~T3 Transfer area H1~H3 Hopper C1~C3 Camera 1 Retention analysis device 2 Arithmetic section 3 Storage section 3a Pre-trained model 4 Display section 5. Communications Department 6. Alert issuing section

Claims

1. A plurality of belt conveyors that transport the objects to be supplied from upstream to downstream; An imaging means is disposed at each transfer section between the plurality of belt conveyors, and captures an image of the object flowing at the transfer section; a first analysis means for acquiring image data captured by the imaging means and analyzing whether the object is flowing normally at the transfer section based on the image data; and an alert issuing means for issuing an alert when it is determined that the object is being supplied at the transfer section as a result of the analysis by the first analysis means, The first analysis means extracts a contour of the object from the video data and analyzes the flow of the object, and when a contour of a vector different from the contour when the object is flowing normally is extracted, the congestion detection system determines that the object is stuck at the transfer area.

2. The image data is input, and information regarding the retention of the object at the transfer section is output as a machine-learned model, The congestion detection system according to claim 1, characterized in that the first analysis means uses the learned model to analyze whether the supplied item is flowing normally at the transfer area.

3. The congestion detection system described in claim 1, further comprising a second analysis means for analyzing whether the supplied goods are flowing normally at the transfer section based on the acquired video data, wherein the analysis by the second analysis means is performed by a method different from that of the first analysis means.

4. A photographing means arranged at each transfer section between the plurality of belt conveyors photographs an image of the state in which the supplied object flows at the transfer section, an image acquisition means for acquiring image data captured by the imaging means; an analysis means for analyzing whether the object is flowing normally at the transfer section based on the image data acquired by the image acquisition means; and an alert issuing means for issuing an alert when it is determined that the object is being supplied at the transfer section as a result of the analysis by the analysis means, The analysis means extracts a contour of the object from the video data to analyze the flow of the object, and when a contour of a vector different from the contour when the object is flowing normally is extracted, the analysis means determines that the object is stuck at the transfer area.

5. A photographing means arranged at each transfer section between the plurality of belt conveyors photographs an image of the state in which the supplied object flows at the transfer section, The computer of the retention analysis device, an image acquisition means for acquiring image data captured by the imaging means; an analysis means for analyzing whether the object is flowing normally at the transfer section based on the image data acquired by the image acquisition means; an alert issuing means for issuing an alert when it is determined that the object is being held up in the transfer section as a result of the analysis by the analysis means; A program that functions as The analysis means extracts a contour of the object from the video data to analyze the flow of the object, and when a contour of a vector different from the contour when the object is flowing normally is extracted, the program determines that the object is stuck at the transfer area.

Images