JP6583947B1 - Level crossing risk determination program and system - Google Patents

Abstract

An object of the present invention is to detect in advance the degree of danger due to contact with a train at a railroad crossing and prevent an accident. In a level crossing risk determination program for determining a level crossing risk, reference image information photographed in a crossing and approach for reference of a train to the level crossing at the time of capturing the reference image information An association level acquisition step for acquiring in advance three or more levels of association between a combination of information and a level of danger for a level crossing for the combination; The information acquisition step refers to the information acquisition step of acquiring the approach information of the train to the railroad crossing at the time of shooting and the association degree acquired in the association degree acquisition step. Based on the image information acquired via the train and the approach information of the train, a discrimination step for discriminating the degree of danger in the crossing is executed on the computer. Characterized in that to. [Selection] Figure 3

Description

  The present invention relates to a level crossing risk determination program and system for determining a risk level within a level crossing.

  Conventionally, an accident caused by contact between a moving body (vehicle, person) and a train at a railroad crossing has been a problem. Although there are many railroad tracks that pass through many railroad crossings, it is possible to detect in advance the degree of danger due to contact between a moving body and a train within the railroad crossing, and to prevent accidents. Desired. It is necessary to perform such a risk automatically and with high accuracy.

Japanese Patent Application No. 10-995

  In the disclosed technique of Patent Document 1 described above, it is described that a learning function of artificial intelligence is provided at a level crossing, but there is no description about determining the degree of risk in a level crossing using artificial intelligence. .

  Therefore, the present invention has been devised in view of the above-described problems, and the object of the present invention is to detect in advance the degree of danger due to contact between a moving body in a railroad crossing and a train, and to prevent an accident from occurring. In order to prevent this, it is an object of the present invention to provide a level crossing risk determination program and system for automatically determining the level of risk in a level crossing using artificial intelligence.

The level crossing risk determination program according to the present invention is a level crossing risk determination program for determining a level crossing risk, the reference image information captured in the crossing, and the level crossing at the time of shooting the reference image information. 3 of the reference approach information of the train to the vehicle, the reference crossing frequency indicating the crossing frequency of the moving body crossing the crossing at the time of photographing the reference image information, and the level of crossing risk for the combination In the association degree acquisition step for acquiring the degree of association in advance at a stage, and when newly determining the risk level at the level crossing, the image information is acquired by newly taking an image in the level crossing, and at the time of the shooting obtaining the cross frequency of the moving body to the railroad crossing, an information acquisition step of acquiring the approach information of the train to the crossing at the shooting time, the linkage A discrimination step for discriminating the degree of risk in the crossing based on the image information obtained through the information acquisition step, the approach information of the train, and the crossing frequency of the moving body with reference to the association degree acquired in the acquisition step. The computer is executed.

  The level crossing risk determination program according to the present invention is a level crossing risk level determination program for determining a level crossing risk level, and the crossing risk information taken within the level crossing and the crossing level at the time when the reference image information is captured. An association level acquisition step for acquiring in advance three or more levels of association between a combination of a reference crossing frequency indicating a crossing frequency of a moving body that crosses the road and a risk level of a level crossing for the combination; When determining the degree, the information acquisition step for acquiring the image information by taking a new image within the level crossing and obtaining the crossing frequency of the moving body to the level crossing at the time of the shooting, and the association degree acquisition step Based on the image information acquired through the information acquisition step and the crossing frequency of the moving body with reference to the association degree acquired in Characterized in that to execute a determining step of determining the risk of the computer.

The level crossing risk determination system according to the present invention is a level crossing risk level determination system for determining the level of crossing risk. The reference image information captured in the crossing and the level crossing at the time of shooting the reference image information. 3 of the reference approach information of the train to the vehicle, the reference crossing frequency indicating the crossing frequency of the moving body crossing the crossing at the time of photographing the reference image information, and the level of crossing risk for the combination When the degree of association is acquired in advance at the level crossing and when the risk level is newly determined at the level crossing, image information is acquired by newly taking an image within the level crossing, and at the time of the shooting obtaining the cross frequency of the moving body to the railroad crossing, an information acquisition unit that acquires the approach information of the train to the crossing of the shooting point, the association degree acquiring means And determining means for determining the degree of risk in the crossing based on the image information acquired by the information acquiring means, train approach information, and the crossing frequency of the moving body. And

  In order to detect in advance the degree of danger due to contact with a train at a level crossing and to prevent an accident, it is possible to automatically determine the degree of danger within a level crossing using artificial intelligence.

It is a block diagram which shows the whole structure of the level crossing risk determination system by which the level crossing risk determination program to which this invention is applied is mounted. It is a figure which shows the specific structural example of a discrimination device. It is a figure for demonstrating the operation | movement in the level crossing risk determination program to which this invention is applied. It is a figure which shows the example which set the 3 or more steps of relevance including the crossing frequency for reference. Furthermore, it is a figure which shows the example which set the 3 or more steps of relevance including reference flow line information. It is a figure which shows the other example which set the 3 or more steps of association degree.

  Hereinafter, a level crossing risk determination program to which the present invention is applied will be described in detail with reference to the drawings.

  FIG. 1 is a block diagram showing an overall configuration of a level crossing risk determination system 1 in which a level crossing risk determination program to which the present invention is applied is installed. The level crossing risk determination system 1 includes an information acquisition unit 9, a determination device 2 connected to the information acquisition unit 9, and a database 3 connected to the determination device 2.

  The information acquisition unit 9 is a device for a person who uses this system to input various commands and information, and specifically includes a keyboard, buttons, a touch panel, a mouse, a switch, and the like. The information acquisition unit 9 is not limited to a device for inputting text information, and may be configured by a device that can detect sound such as a microphone and convert it into text information. The information acquisition unit 9 may be configured as an imaging device that can capture an image such as a camera. The information acquisition unit 9 may be configured by a scanner having a function of recognizing a character string from a paper medium document. Moreover, the information acquisition part 9 may be integrated with the discrimination | determination apparatus 2 mentioned later. The information acquisition unit 9 outputs the detected information to the determination device 2.

  The database 3 stores information related to the risk level in the level crossing, such as accidents that have occurred at the level crossing before, or cases that did not lead to an accident but had a high level of risk. The database 3 includes reference image information that is actually captured by a camera that actually constitutes the information acquisition unit 9, reference access information of a train to the railroad crossing at the time when the reference image information is captured, and a moving object that crosses the railroad crossing. The reference crossing frequency indicating the crossing frequency, the flow line for reference that extracts the flow line of the moving body crossing the railroad crossing (including the time-series change of the flow line), and the like are accumulated.

  The discriminating device 2 is composed of electronic devices such as a personal computer (PC), for example. In addition to the PC, the discriminating device 2 can be embodied in any other electronic device such as a mobile phone, a smartphone, a tablet terminal, and a wearable terminal. It may be made. The user can discriminate the degree of danger at the crossing by obtaining a search solution by the discriminating device 2.

  FIG. 2 shows a specific configuration example of the discrimination device 2. The discrimination device 2 performs wired communication or wireless communication with a control unit 24 for controlling the discrimination device 2 as a whole, and an operation unit 25 for inputting various control commands via operation buttons, a keyboard, and the like. A communication unit 26 for making a determination, a determination unit 27 for making various determinations, and a storage unit 28 for storing a program for performing a search to be executed, represented by a hard disk or the like, are connected to the internal bus 21, respectively. . Further, the internal bus 21 is connected to a display unit 23 as a monitor for actually displaying information.

  The control unit 24 is a so-called central control unit for controlling each component mounted in the determination device 2 by transmitting a control signal via the internal bus 21. Further, the control unit 24 transmits various control commands via the internal bus 21 in accordance with an operation via the operation unit 25.

  The operation unit 25 is embodied by a keyboard or a touch panel, and an execution command for executing a program is input from the user. When the execution command is input from the user, the operation unit 25 notifies the control unit 24 of this. Upon receiving this notification, the control unit 24 executes a desired processing operation in cooperation with each component including the determination unit 27. The operation unit 25 may be embodied as the information acquisition unit 9 described above.

  The determination unit 27 is responsible for various determinations regarding the level of danger of a crossing. The determination unit 27 reads various information stored in the storage unit 28 as necessary information and various information stored in the database 3 when executing the estimation operation. This determination unit 27 may be controlled by artificial intelligence. This artificial intelligence may be based on any known artificial intelligence technology.

  The display unit 23 includes a graphic controller that creates a display image based on control by the control unit 24. The display unit 23 is realized by, for example, a liquid crystal display (LCD).

  When the storage unit 28 is composed of a hard disk, based on the control by the control unit 24, predetermined information is written to each address and is read out as necessary. The storage unit 28 stores a program for executing the present invention. This program is read by the control unit 24 and executed.

  The operation in the level crossing risk determination system 1 having the above-described configuration will be described.

  In the level crossing risk determination system 1, for example, as shown in FIG. 3, it is assumed that a combination of reference image information and reference approach information is formed. Reference image information is an image of various parts of a level crossing taken by a camera installed at the level crossing. For example, a vehicle or a person or other moving body is crossed in a state where the circuit breaker is not down. It is an image that is running, moving with a moving body crowded with one of the circuit breakers down, or a wheelchair with all the circuit breakers down There are various images such as an image of a train approaching a railroad crossing even though the person on the railroad is on the railroad crossing.

  In the example of FIG. 3, for example, the reference image information includes reference image information P <b> 11 to P <b> 13 photographed at each place or each time series with a camera installed at the crossing, and the reference approach information is a train. It is assumed that the time when the vehicle approaches the railroad crossing is “down train, 30 seconds later”, “up train, 45 seconds later”, and the like.

  As input data, such reference image information and reference approach information are arranged side by side. The intermediate node shown in FIG. 3 is a combination of the reference image information as the input data and the reference approach information. Each intermediate node is further connected to an output. In this output, the degree of danger in the crossing as an output solution is displayed as a percentage.

  Each combination (intermediate node) of the reference image information and the reference approach information is associated with each other through the three or more levels of relevance with respect to the risk level in the crossing as the output solution. The reference image information and the reference approach information are arranged on the left side through the association degree, and the respective risk degrees are arranged on the right side through the association degree. The degree of association indicates the degree of risk and the degree of relevance of the reference image information and the reference approach information arranged on the left side. In other words, the degree of association is an index indicating to which risk level each reference image information and reference approach information are likely to be linked, and is most determined from the reference image information and the reference approach information. It shows the accuracy in selecting a certain risk level. In the example of FIG. 3, w13 to w22 are shown as the association degrees. These w13 to w22 are shown in 10 stages as shown in Table 1 below. The closer to 10 points, the higher the degree of relevance between each combination as an intermediate node and the level of danger in the level crossing as an output. On the contrary, the closer to one point, the lower the degree of association between each combination as an intermediate node and the degree of danger in the level crossing as an output.

  The discriminating apparatus 2 obtains in advance three or more levels of association w13 to w22 shown in FIG. In other words, the discriminating device 2 accumulates reference image information, reference approach information, and data indicating how much the risk level is in that case, and analyzes these in determining the actual risk level. By analyzing, the relevance shown in FIG. 3 is created.

  For example, it is assumed that the reference image information P11 is an image in which a moving body such as a vehicle or a person crosses in a sparse state in a state where the circuit breaker is not descended. When the reference approach information at the time of this image capture is that the train will come “up train, 45 seconds later”, whether or not an accident occurred in the previous data, and an accident occurred even if it did not happen It is extracted whether or not the degree of danger is not strange. These data may be extracted from past accident data stored at a railway company or each station, or from data of a scene that has been curious. The digitization of the risk level may be obtained by visualizing the reference image information by a plurality of people and counting questionnaire surveys on the risk level.

  This analysis and analysis may be performed by artificial intelligence. In such a case, for example, in the case of the reference image information P11 and “up train, 45 seconds later”, whether or not an accident has actually occurred or whether or not the accident has occurred is high. Analyzing this from past data. The higher the number of cases in which an accident has occurred, the higher the degree of association that leads to high-risk output, and the lower the number of cases in which an accident has occurred, the higher the degree of association that leads to low-risk output. For example, in the example of the intermediate node 61a, the output is linked to an output of 90% risk and 30% risk. However, since the risk is extremely high from the previous case, w13 linkage that leads to 90% risk. The degree of association of w14, which leads to a degree of risk of 30%, is set to 2 points.

  Further, the association degree shown in FIG. 3 may be composed of nodes of a neural network in artificial intelligence. That is, the weighting coefficient for the output of the node of this neural network corresponds to the above-mentioned relevance. Moreover, it is not limited to a neural network, and may be composed of all decision-making factors that constitute artificial intelligence.

  In the example of the association degree shown in FIG. 3, the node 61b is a node that is a combination of the reference approach information “upward train, 2 minutes later” with respect to the reference image information P11, and has an association degree of 60% risk. The relevance of w15 and risk 0% is w16. The node 61c is a node having a combination of the reference approach information “down train, 30 seconds later” and “down train, 15 seconds later” with respect to the reference image information P12. The relevance with a risk level of 70% is w18.

  Such association is the learned data in terms of artificial intelligence. After creating such learned data, when determining a new risk level at a new level crossing, the risk level is determined using the learned data described above. In such a case, the image information is newly acquired and the approach information is acquired.

  The newly acquired image information is taken by the camera by the information acquisition unit 9 described above. This photographing is the same as the railroad crossing photographed for obtaining the reference image information described above. Further, the shooting conditions are not required until the shooting conditions (shooting angle, angle of view, resolution) for obtaining the above-described reference image information are all the same.

  The approach information may be acquired directly from train operation data managed by a railway company. The approach information may be acquired based on one or more of train schedule information and train delay information. The actual number of minutes a train arrives at a certain level crossing is based on the current time, train schedule, delay information, climate, weather, and the like. From this information, it is calculated how many minutes the train will actually arrive at the railroad crossing. Further, the approach information may be acquired based on whether or not the train has passed the 50m point before the railroad crossing. In other words, if the sensor detects that the train has passed 50m before the railroad crossing, it will determine that the train is approaching, otherwise it will determine that the train is not approaching, This may be included in the approach information. The approach information acquisition method is the same when acquiring the reference approach information described above.

  Based on the image information newly acquired in this way and the approach information of the train, the degree of risk at the time when the image information and the approach information are actually acquired is obtained. In such a case, the association degree shown in FIG. 3 (Table 1) acquired in advance is referred to. For example, if the newly acquired image information is the same as or similar to P12 and the approach information obtained at the time of acquisition is “down train, 15 seconds later”, the degree of association The node 61d is associated with the “risk degree 60%” by w19 and the “risk degree 70%” by the association degree w20. In such a case, “risk level 60%” having a higher relevance is selected as the optimal solution. However, it is not essential to select a solution with the highest degree of association as an optimal solution, and “risk degree 70%” where the association itself is recognized although the association degree is low may be selected as the optimum solution. In addition to this, it is of course possible to select an output solution that is not connected to an arrow, and any other solution may be selected as long as it is based on the degree of association.

  Table 2 below shows examples of the association degrees w1 to w12 extending from the input.

  The intermediate node 61 may be selected based on the association degrees w1 to w12 extending from this input. That is, the weighting in the selection of the intermediate node 61 may be increased as the association degrees w1 to w12 are increased. However, the association degrees w1 to w12 may all be the same value, and all the weights in the selection of the intermediate node 61 may be the same.

  In FIG. 4, in addition to the reference image information and the reference approach information described above, three or more levels of relevance of the combination of the reference crossing frequency and the risk level in the crossing for the combination are set. An example is shown.

  The reference crossing frequency may be quantified by the density per unit area of the moving body that crosses the level crossing, or may be quantified through the number of moving bodies that cross the level crossing per unit time. Good. At this time, the crossing speed of the moving body may be measured and included as a parameter. Since the reference crossing frequency changes depending on the time, the reference crossing frequency may be associated with the time point when the reference image information is captured and the time point when the reference approach information is acquired.

  In this case, as shown in FIG. 4, the association degree is expressed as a set of combinations of the reference image information, the reference approach information, and the reference crossing frequency as nodes 61a to 61e of the intermediate nodes as described above. Will be.

  For example, in FIG. 4, the node 61c has the reference image information P12 with the association degree w3, the reference approach information “down train, 30 seconds later” has the association degree w7, and the “crossing frequency β ( For example, the average moving amount of the moving body per minute is 6) ”is associated with the association degree w11. Similarly, the node 61e has a relevance of w5 for the reference image information P13, a relevance of w8 for the reference approach information “down train, 30 seconds later”, and the “crossing frequency α (for example, unit area) as the reference crossing frequency. The number of hitting moving bodies is 3), etc.

  Similarly, when such association is set, based on newly acquired image information, train approach information, and the crossing frequency to the railroad crossing at the time of shooting, the image information is actually newly added. And the degree of danger at the time when the approach information is acquired. As for the crossing frequency here, if the shooting time (what day, what time, what minute) of the newly acquired image information is known, the one at the time of shooting is read out from the reference crossing frequency already recorded. Or the frequency of crossing may be measured each time. The crossing frequency may be obtained, for example, by measuring the crossing frequency for 3 minutes in advance.

  In obtaining this risk level, the association degree shown in FIG. 4 acquired in advance is referred to. For example, when the acquired image is the same as or similar to the reference image information P12, the access information “down train, 30 seconds later”, and the acquired crossing frequency is the crossing frequency β, the combination is associated with the node 61c. In this node 61c, the degree of risk 30% is associated with the association degree w17, and the degree of risk 70% is associated with the association degree w18. As a result of such association, the risk level at the time when the image information and the approach information are actually newly acquired is obtained based on w17 and w18.

  In FIG. 5, in addition to the above-described reference image information and reference approach information, a combination of reference flow line information and a degree of relevance of three or more levels of the risk in the crossing for the combination are set. An example is shown.

  The reference flow line information is extracted from the moving line of the moving body in the railroad crossing in a certain time zone (for example, it may be a temporary point or a time width from what hour to what hour to what minute). Is. In extracting the reference flow line information, the moving body flow line may be detected by a well-known means by performing image analysis on the above-described reference image information. The pattern of the moving line of the moving body is patterned and stored on the image by a well-known means through a vector or a diagram.

  In such a case, as shown in FIG. 5, the association degree is obtained by combining a set of combinations of reference image information, reference approach information, and reference flow line information as intermediate nodes 61 a to 61 e as described above. It will be expressed.

  For example, in FIG. 5, the node 61c has the reference image information P12 with the association degree w3, the reference approach information “down train, 30 seconds later” with the association degree w7, and the passenger flow line as the reference flow line information. W is associated with the association degree w11. Similarly, in the node 61e, the reference image information P13 has the association degree w5, the reference approach information “down train, 30 seconds later” has the association degree w8, and the passenger's flow line V as the reference flow line information is the association degree. Linked with w10.

  Similarly, when such a degree of association is set, the newly acquired image information, the train approach information, and the flow pattern of the moving object at the time when the image information and the approach information are acquired are further acquired. To do. The extraction of the movement line pattern of the moving object is the same as the extraction method of the reference passenger movement line information.

  In obtaining this risk level, the association degree shown in FIG. 5 acquired in advance is referred to. For example, when the acquired image is the same as or similar to the reference image information P12, the acquired approach information is “down train, 30 seconds later”, and the acquired movement line of the moving object is the flow line W, The node 61c is associated with the combination, and the node 61c is associated with the association degree w17 with a risk level of 30% and with the association degree w18. As a result of such association, the risk level at the time when the image information and the approach information are actually newly acquired is obtained based on w17 and w18.

  FIG. 6 shows an example in which three or more levels of association are set between the combination of the reference image information and the reference crossing information described above and the risk level in the crossing for the combination.

  As input data, such reference image information and reference crossing information are arranged side by side. The intermediate node shown in FIG. 6 is a combination of reference image information and reference crossing information as such input data.

  The discriminating apparatus 2 acquires in advance the three or more levels of association w13 to w22 shown in FIG. In other words, the discriminating device 2 accumulates reference image information, reference crossing frequency, and data indicating the degree of risk in that case, and analyzes these in determining the actual risk level. By analyzing, the association shown in FIG. 6 is created.

  In the example of the association degree shown in FIG. 6, the node 61b is a node having a combination of the reference crossing frequency “crossing frequency α” with respect to the reference image information P11. The association degree of 0% is w16.

  Similarly, when such a degree of association is set, based on the newly acquired image information and the crossing frequency to the level crossing at the time of shooting, the image information and the approach information are actually acquired. Find the degree of danger at the time.

  In obtaining this risk level, the association level shown in FIG. 6 acquired in advance is referred to. For example, when the acquired image is the same as or similar to the reference image information P11 and the acquired crossing frequency is the crossing frequency α, the combination is associated with the node 61b, and the degree of association with the output at the node 61b is The risk will be determined.

  In addition, in the above-described association degree, the case where it is formed with respect to any combination of the reference flow information and the reference crossing frequency in addition to the reference image information and the reference approach information has been described as an example. It is not limited to this. In addition to the reference image information and the reference approach information, the association degree may be associated with one or more combinations of the reference flow line information and the reference crossing frequency.

  In the above-described association degree, the association degree may be associated with a combination of reference flow line information in addition to the reference image information and the reference crossing frequency. Further, the determination may be made in consideration of audio information in addition to the reference image information. The sound information is sound detected by a microphone or the like. This voice information may be detected in the same manner as when the reference image is acquired and may be associated with the association degree as an independent input parameter as the reference voice information. That is, in addition to the reference image information and the reference approach information, the association degree may be associated with a combination of the reference audio information. In such a case, voice information is acquired with a microphone in the determination of the new risk level, and the risk level is determined with reference to the association level.

  In the above-mentioned relevance, the relevance is expressed by a 10-level evaluation, but it is not limited to this, and it may be expressed by a relevance of 3 or more levels. For example, 100 steps or 1000 steps may be used. On the other hand, this degree of association does not include two levels, that is, whether or not they are associated with each other, and those expressed with either 1 or 0.

  In addition, although the danger level demonstrated taking the case where it described by the percentage of 0 to 100% as an example, it is not limited to this. This risk level may be expressed in two stages, for example, “high risk” and “low risk”. In this case, when forming the relevance, the combination of the reference image information and the reference approach information, etc. is classified as “high risk” or “danger” when analyzing the judgment example of whether the risk is high or not. It will be analyzed and discriminated in association with the result of “low performance”. Of course, these operations may be replaced by artificial intelligence.

  According to the present invention having the above-described configuration, it is possible to easily determine the degree of risk at a level crossing with little effort without requiring special skill. Further, according to the present invention, it is possible to determine the degree of danger at this level crossing with higher accuracy than humans can perform. Furthermore, by configuring the above-mentioned association degree with artificial intelligence (such as a neural network), it is possible to further improve the discrimination accuracy by learning this.

  In addition, the present invention is characterized in that an optimum physical property and a generation mechanism are searched through the association degree set in three or more stages. The degree of association can be described by a numerical value of, for example, 0 to 100% in addition to the above five levels, but is not limited to this, and any level can be described as long as it can be described by a numerical value of three or more levels. It may be configured.

  By searching for the most probable risk based on the degree of association represented by numerical values of three or more levels in this way, in the situation where there are a plurality of candidates for the possibility of a high risk at the level crossing, It is also possible to search and display in descending order. If it is possible to display to the user in the order of the degree of association in this way, it is possible to preferentially display a more certain danger level, and it is possible to call attention due to an increase in the risk level.

  In particular, when the degree of danger is higher, it is preceded by processing operations such as alerting the moving body through voice, etc., or notifying the train entering the railroad crossing that the danger is high and voluntarily stopping it. This can be done and can keep passengers safe. In addition, according to the present invention, since the detection accuracy of the danger level is high, it is only necessary to perform the operation of stopping the stop train only in the case where it is really necessary without stopping the train unnecessarily, and thus adversely affects smooth traffic. This can also be prevented.

  In addition to this, according to the present invention, it is possible to make a determination without overlooking a discrimination result of an extremely low output such as an association degree of 1%. Even a discrimination result with a very low degree of association is connected as a small sign, and alerts the user that it may be useful as a discrimination result once every tens or hundreds of times. be able to.

  Furthermore, according to the present invention, there is an advantage that a search policy can be determined by a method of setting a threshold value by performing a search based on such three or more levels of association. If the threshold value is lowered, it is possible to pick up without omission even if the above-mentioned association degree is 1%, but there is a low possibility that a more appropriate discrimination result can be suitably detected, and a lot of noise may be picked up. is there. On the other hand, if the threshold value is increased, it is likely that the optimum risk level can be detected with a high probability. However, although the degree of association is usually low and it is slewed, it is suitable to appear once every tens or hundreds of times. Sometimes the solution is overlooked. It is possible to decide which to place importance on the basis of the idea on the user side and the system side, but it is possible to increase the degree of freedom in selecting points to place such emphasis.

  Furthermore, in the present invention, the association degree described above may be updated. This update may reflect information provided via a public communication network such as the Internet. In addition, when new knowledge is discovered about the relationship between input parameters and output solutions (risk levels) based on camera images uniquely captured by electric railway companies and stations, and access information uniquely acquired Depending on the knowledge, the association degree is increased or decreased.

  That is, this update is equivalent to learning in terms of artificial intelligence. It is a learning act because new data is acquired and reflected in the learned data.

  In addition, the process of first creating a learned model and the above-described updating may use not only supervised learning but also unsupervised learning, deep learning, reinforcement learning, and the like. In the case of unsupervised learning, instead of reading and learning a data set of input data and output data, information corresponding to input data (reference image information and reference approach information, reference flow line information, Reference crossing frequency or the like) may be read and learned, and the degree of association associated with the output data may be self-generated from there.

  In addition to information based on information that can be obtained from the public communication network, this relevance update is artificially performed by the system side or user side based on the contents of research data, papers, conference presentations, newspaper articles, books, etc. by experts. Alternatively, it may be updated automatically. Artificial intelligence may be used in these update processes.

DESCRIPTION OF SYMBOLS 1 Level crossing risk discrimination system 2 Discriminating device 21 Internal bus 23 Display part 24 Control part 25 Operation part 26 Communication part 27 Estimation part 28 Storage part 61 Node

Claims (8)

In a level crossing risk determination program for determining a level crossing risk,
Reference image information taken at the level crossing, train reference approach information to the level crossing at the time of shooting the reference image information, and a moving object that crosses the level crossing at the time of shooting of the reference image information An association degree acquisition step of acquiring in advance three or more degrees of association between a combination of a reference crossing frequency indicating a crossing frequency and a level of crossing risk for the combination;
When a new risk level is determined at the level crossing, image information is acquired by taking a new image within the level crossing, and the crossing frequency of the moving body to the level crossing at the time of the shooting is obtained. An information acquisition step for acquiring train approach information to the railroad crossing at
Discrimination that determines the degree of risk in the level crossing with reference to the association degree acquired in the association degree acquisition step, based on the image information acquired through the information acquisition step, the approach information of the train, and the crossing frequency of the moving object A program for determining a level crossing risk, which causes a computer to execute steps. In a level crossing risk determination program for determining a level crossing risk,
A combination of reference image information taken in a crossing and a reference crossing frequency indicating a crossing frequency of a moving body that crosses the crossing at the time of shooting the reference image information, and a risk of crossing for the combination An association degree acquisition step for acquiring the degree of association of three or more stages in advance;
An information acquisition step for acquiring image information by newly taking an image in a crossing when newly determining the degree of risk at the crossing, and obtaining a crossing frequency of the moving body to the crossing at the time of shooting ,
A discrimination step for discriminating the degree of risk in the crossing based on the image information obtained through the information acquisition step and the crossing frequency of the moving body with reference to the association degree obtained in the association degree obtaining step is performed on the computer. A level crossing risk determination program characterized by being executed. In the relevance level acquisition step, three or more levels of relevance are acquired in advance: a combination of reference flow line information obtained by extracting a flow line of a moving body that crosses the crossing and a risk level in the crossing corresponding to the combination. And
In the information acquisition step, when newly determining the risk level at the crossing, the flow line information obtained by extracting the flow line of the moving body is further acquired,
The level crossing risk determination program according to claim 1 or 2, wherein, in the determination step, the risk level in the level crossing is determined based on the flow line information acquired through the information acquisition step. In a level crossing risk determination program for determining a level crossing risk,
A combination of reference image information taken at the level crossing, the reference reference information of the train to the level crossing at the time of shooting the reference image information, and the level of risk of the level crossing for the combination in three or more levels An association degree obtaining step for obtaining the degree in advance;
An information acquisition step of acquiring image information by newly taking an image in the level crossing and acquiring approach information of a train to the level crossing at the time of the imaging when newly determining the risk level at the level crossing; ,
A determination step for determining the degree of danger in the crossing based on the image information acquired through the information acquisition step and the approach information of the train with reference to the association degree acquired in the association degree acquisition step;
In the association degree acquisition step, reference approach information is acquired based on any one or more of train diagram information and train delay information at the time of photographing the reference image information,
In the information acquisition step, when the risk level is newly determined at the crossing, the approach information is acquired based on one or more of the train diagram information and the train delay information acquired at the time of the photographing. A level crossing risk determination program that is executed by a computer. The level crossing risk determination program according to any one of claims 1 to 4, wherein, in the association degree acquisition step, the association degree constituted by nodes of a neural network in artificial intelligence is acquired in advance. In a level crossing risk determination system for determining a level crossing risk,
Reference image information taken at the level crossing, train reference approach information to the level crossing at the time of shooting the reference image information, and a moving object that crosses the level crossing at the time of shooting of the reference image information A degree-of-association acquisition means for acquiring in advance three or more levels of association between a combination of a reference crossing frequency indicating a crossing frequency and a level of crossing risk for the combination;
When a new risk level is determined at the level crossing, image information is acquired by taking a new image within the level crossing, and the crossing frequency of the moving body to the level crossing at the time of the shooting is obtained. Information acquisition means for acquiring approach information of the train to the railroad crossing at
Discrimination means for discriminating the degree of risk in the railroad crossing based on the image information obtained by the information obtaining means, the approach information of the train, and the crossing frequency of the moving body, with reference to the association degree obtained by the association degree obtaining means; A level crossing risk determination system characterized by comprising: In a level crossing risk determination system for determining a level crossing risk,
  A combination of reference image information taken in a crossing and a reference crossing frequency indicating a crossing frequency of a moving body that crosses the crossing at the time of shooting the reference image information, and a risk of crossing for the combination A degree-of-association acquisition means for acquiring the degree of association in three or more stages in advance;
  When newly determining the risk level at the crossing, information acquisition means for acquiring image information by newly taking an image in the crossing and obtaining a crossing frequency of the moving body to the crossing at the time of the shooting, ,
  A determining unit that refers to the association degree acquired by the association degree acquiring unit and determines the risk level in the crossing based on the image information acquired by the information acquiring unit and the crossing frequency of the moving object;
  A system for determining the level of danger at a level crossing. In a level crossing risk determination system for determining a level crossing risk,
A combination of reference image information taken at the level crossing, the reference reference information of the train to the level crossing at the time of shooting the reference image information, and the level of risk of the level crossing for the combination in three or more levels An association degree acquisition means for acquiring the degree in advance;
An information acquisition means for acquiring image information by newly taking an image in a crossing and acquiring information on approach of a train to the crossing at the time of shooting when newly determining the risk level at the crossing; ,
With reference to the association degree acquired in the association degree acquisition means, and a determination means for determining the risk level in the crossing based on the image information acquired by the information acquisition means and train approach information,
The association degree obtaining means obtains reference approach information based on one or more of train diagram information and train delay information at the time of photographing the reference image information,
The information acquisition means acquires the approach information based on one or more of the train diagram information and the train delay information acquired at the time of photographing when the risk level is newly determined at the crossing.
A system for determining the level of danger at a level crossing .

Images