Tunnel Image Recognition-Based Automatic Emergency Communication Interlocking System

Abstract

The present invention relates to a system for automatically recognizing accident situations within a tunnel and automatically controlling emergency communication facilities in conjunction therewith. Specifically, it relates to a tunnel image recognition-based emergency communication automatic interlocking system that analyzes video and measurement data collected from multiple tunnels at an integrated control server, calculates a risk index, and automatically performs section-by-section emergency broadcasting, emergency telephone activation, notification to external agencies, and transmission network control.

Description

Technology Field

[0001] The present invention relates to a system for automatically recognizing accident situations within a tunnel and automatically controlling emergency communication facilities in conjunction therewith. Specifically, it relates to a tunnel image recognition-based emergency communication automatic interlocking system that analyzes video and measurement data collected from multiple tunnels at an integrated control server, calculates a risk index, and automatically performs section-by-section emergency broadcasting, emergency telephone activation, notification to external agencies, and transmission network control. Background Technology

[0002] Tunnels are specialized transportation facilities characterized by enclosed spatial structures and limited access points. Unlike ordinary roads, they possess environmental characteristics that make the spread of danger rapid and evacuation difficult in the event of an accident. In particular, accidents involving vehicle stoppages, driving against the flow of traffic, fires, smoke diffusion, and the increase in hazardous gases can affect a large number of vehicles within a short period, making the speed and accuracy of initial response critical.

[0003] Conventional tunnel management systems are primarily operated in the following ways.

[0004] First, the general method involves filming the inside of the tunnel via multiple CCTVs, with control personnel monitoring the footage to determine if an accident has occurred and then manually making emergency broadcasts or notifying the outside. However, this structure can lead to delays in initial response times due to issues such as delayed awareness, errors in judgment, and reduced concentration caused by shift work.

[0005] Second, while some systems are configured to automatically recognize vehicle stoppages or fires by applying video analysis technology, they are limited to determining the occurrence of simple events and often fail to sufficiently incorporate quantitative analysis structures, such as the simultaneous occurrence of multiple events or risk assessment based on rates of change. For instance, risk index calculation structures that comprehensively reflect temporal change characteristics—such as the rate of rapid vehicle speed decrease, the rate of increase in smoke area, and the rate of change in flame brightness—are currently inadequate.

[0006] Third, in conventional technology, video detection systems and emergency communication facilities (emergency broadcasts, emergency telephones, etc.) are often operated individually or remain in a simple interlocking structure. Consequently, it is common practice to perform simultaneous broadcasting across the entire tunnel rather than selective control limited to the accident section. This can cause alarms to be triggered in sections unrelated to the accident site, potentially leading to driver confusion and raising concerns about overreaction.

[0007] Fourth, due to the enclosed structure of tunnels, smoke and harmful gases have the characteristic of rapidly spreading in specific directions depending on the ventilation direction; however, conventional systems often fail to actively reflect the operating direction of ventilation equipment or air flow information in risk assessment. Consequently, there is a problem in that preemptive response to the actual sections where risk spread is expected is not sufficiently carried out.

[0008] Fifth, when an accident occurs, video transmission volume and communication traffic surge. However, conventional transmission network structures often fail to clearly separate emergency communication packets from general data packets or are limited to simple priority settings, which limits the stable transmission of actually necessary emergency communications to a certain degree. In particular, there are cases where a structure that automatically switches to a backup path in the event of transmission delay is not sufficiently provided.

[0009] Sixth, although tunnels have a structure where the direction of vehicle travel is clearly distinguished and require different guidance for vehicles ahead and behind in the event of an accident, conventional systems do not sufficiently include a differential response structure that reflects the direction of travel.

[0010] Seventh, when managing multiple tunnels in an integrated manner, the lack of a system to comprehensively manage structural information, section mapping information, and equipment location information for each tunnel and reflect them in risk assessment often results in fragmented control centered on individual equipment.

[0011] As such, although conventional technology consists of separate image detection, communication control, and integrated server operation, there were limitations in implementing them into an integrated automatic response system that combines a quantitative risk calculation structure reflecting the structural and environmental characteristics of the tunnel with segment-unit selective response and communication network structure control.

[0012] Therefore, a dedicated automatic response system for tunnels is required that calculates a risk index by comprehensively reflecting the quantitative rate of change of vehicle movement status, smoke diffusion characteristics, ventilation direction information, tunnel section information, and communication network status information, and ensures communication stability by automatically executing section-unit emergency communication based on the risk index while simultaneously controlling the transmission network structure. Prior art literature

[0013] (Patent Document 0001) KR 10-2899844 B1(2025.12.09) The problem to be solved

[0014] The present invention aims to provide a tunnel video recognition-based emergency communication automatic interlocking system capable of automatically recognizing accident situations occurring within a tunnel, such as vehicle stopping, driving in reverse, fire, and smoke spread, based on CCTV video and measurement data, calculating a quantitative risk index, and automatically interlocking and controlling emergency communication facilities.

[0015] Furthermore, the present invention aims to provide an automatic response structure optimized for tunnel-specific environments by determining complex risks by comprehensively reflecting the rate of change in vehicle movement status, the rate of increase in smoke, and the rate of rise in harmful gas concentration, and by performing selective section response by considering tunnel section information, the direction of vehicle movement, and the direction of ventilation.

[0016] In addition, the present invention aims to improve the reliability of emergency communication and system stability by configuring an independent transmission path for video of the accident section, performing automatic path switching when communication delay occurs, and activating adjacent section equipment as an alternative in the event of equipment failure. means of solving the problem

[0017] An embodiment of the present invention relates to a tunnel image recognition-based emergency communication automatic interlocking system that recognizes events such as vehicle stopping, driving in reverse, fire, or smoke spread occurring within a tunnel based on image and automatically interlocks emergency communication facilities, comprising: a plurality of cameras installed inside or outside the tunnel to capture images of the tunnel situation and transmit the said images to an emergency response control terminal; a risk analysis unit that analyzes image data received from the cameras to calculate event type information, event occurrence coordinate information, reliability value, risk index, and risk level information, and calculates the calculated risk index and risk level information; and a section management unit that receives the risk index, risk level information, and event occurrence coordinate information transmitted from the risk analysis unit, divides the tunnel into a plurality of sections to map the event occurrence location to a section identifier, sets response target sections based on the risk level and expected spread range, and transmits the set response target section identifier list to a communication execution unit. A communication execution unit that receives a list of response target section identifiers and risk level information transmitted from a section management unit, generates control data packets based on a response scenario, automatically activates emergency terminals mapped to response target sections based on the generated control data packets, establishes a call session to perform automatic dialing to a set public institution, and transmits notification data including event type information, risk level information, and event occurrence coordinate information to an external public institution system or an external traffic system; a network control unit that receives network control request information and communication network resource status information transmitted from the communication execution unit, and minimizes transmission delay of emergency control data by adjusting at least one of transmission bandwidth, transmission priority, or transmission path based on the risk level and required transmission conditions;An environment correction unit that calculates an estimated spreading distance and estimated spreading time by combining risk analysis results and environmental sensor information, transmits the calculated estimated spreading distance and estimated spreading time to a section management unit to reflect them in the setting of response target sections, calculates a correction risk index and reflects it in the control judgment of a communication execution unit, and transmits an entry restriction control signal to at least one of a tunnel entrance control facility and an external traffic system when the correction risk index exceeds a set standard; wherein the emergency terminal is a communication terminal installed inside the tunnel, configured to automatically open a call channel based on an activation signal received from the communication execution unit, transmit status information including terminal identifier information and installation location coordinate information to an external traffic system or a set public institution, and transmit and receive voice data during the call session maintenance time; is provided.

[0018] The risk analysis unit calculates a risk index by normalizing and weighting at least one of the vehicle speed reduction rate, stopping duration, smoke area increase rate, or flame brightness change rate, and classifies the calculated risk index into multiple stages to determine a warning stage or an emergency stage.

[0019] In addition, the section management unit is characterized by mapping event occurrence location information to a tunnel distance coordinate system, analyzing vehicle travel direction vectors to distinguish between the forward and rear sections, and being configured to additionally include adjacent sections or multiple forward sections in the corresponding target sections according to the risk level.

[0020] In addition, the section management unit dynamically expands the corresponding target section based on the estimated spreading distance and estimated spreading time, and calculates multiple section identifiers included in the estimated spreading range and transmits them to the communication execution unit.

[0021] Additionally, the communication execution unit (300) is characterized by being configured to sequentially execute emergency broadcast transmission, automatic activation of emergency terminals, and notification to an external public institution system (30) when the danger level is classified as an emergency level, and to prioritize the execution of only the emergency broadcast when the danger level is classified as a warning level.

[0022] In addition, the emergency terminal is characterized by being configured to automatically make a call to the control center connection number based on an activation signal received from the communication execution unit, and to terminate the session when the call session maintenance time elapses or a call termination signal is received.

[0023] Additionally, the network control unit includes: a priority control module that resets the transmission priority by placing emergency control data packets in the upper queue among multiple transmission queues and moving general data packets to the lower queue based on data type code, required bandwidth value, priority level, and maximum allowable delay time information included in network control request information received from the communication execution unit (300); a bandwidth control module that allocates priority bandwidth to emergency control data by comparing the required bandwidth value with communication network resource status information received from the status monitoring module to limit the transmission rate of general traffic, redistribute transmission queue resources, or reset virtual network resources; a path switching module that compares the average delay time and packet loss rate of the main transmission path with the maximum allowable delay time or reference loss rate, and switches the transmission session to a reserve transmission path if the average delay time or packet loss rate exceeds the reference; a status monitoring module that generates communication network resource status information by collecting transmission bandwidth usage, average delay time, packet loss rate, and port occupancy information from network equipment and transmits it to the priority control module, bandwidth control module, and path switching module; and an emergency channel creation module that creates a dedicated virtual transmission channel separated from existing communication traffic and sets the emergency control data packets to be transmitted through the dedicated channel when the danger stage is determined to be an emergency stage. A predictive resource reservation module that calculates an estimated data transmission volume corresponding to additional broadcast transmissions or external notifications expected during the spread period based on the estimated spread distance and estimated spread time, and preemptively reserves bandwidth corresponding to the calculated transmission volume to separate it from general traffic;An independent verification module that compares primary communication network status information received from a status monitoring module (440) with secondary communication network status information calculated through multiple actual measurements on the same transmission path, calculates an average value, calculates a standard deviation, and removes outliers, and determines whether there is a communication network abnormality if the difference between the two pieces of information exceeds a set standard, and transmits a switching signal to a path switching module and transmits a network reliability degradation index to a risk analysis unit when a communication network abnormality is confirmed; and a section weighting adjustment module that receives section-specific accident history information and traffic density information to calculate section-specific weighting coefficients, reflects the calculated weighting coefficients in the priority level and bandwidth allocation ratio of emergency control data packets, and transmits a section risk accumulation index to a risk analysis unit to reflect it in the recalculation of the risk index when an increase in repetitive accident history is detected.

[0024] Additionally, the environment correction unit calculates the risk diffusion path, estimated diffusion distance, and estimated diffusion time by combining event occurrence location information and event intensity information transmitted from the risk analysis unit with tunnel internal wind speed and airflow direction information received from the environment sensor, transmits the calculated estimated diffusion distance and estimated diffusion time to the section management unit to reflect them in the setting of the target section for response, and transmits them to the prediction resource reservation module to reflect them in the calculation of the estimated transmission resource; the acoustic correction module adjusts the emergency broadcast sound pressure level and output amplification rate based on tunnel cross-section structure information and reflection coefficient information stored in the structure storage module and background noise level information received from the environment sensor, and transmits the adjusted acoustic output setting value to the communication execution unit to reflect it in the broadcast transmission conditions; and the risk correction module calculates a risk index correction coefficient by combining the estimated diffusion distance and estimated diffusion time transmitted from the diffusion prediction module with tunnel internal traffic density information and ventilation status information, recalculates the corrected risk index by applying the correction coefficient to the existing risk index with weights, and transmits the recalculated corrected risk index to the communication execution unit to reflect it in the control stage judgment. It includes an entry interlocking module that determines whether to restrict entry into the tunnel by combining the expected spread distance transmitted from the spread prediction module and the corrected risk index transmitted from the risk correction module, and if the corrected risk index exceeds the set criteria, transmits an entry restriction control signal to the tunnel entrance control facility or an external traffic system and transmits the control result information to the communication execution unit. Effects of the invention

[0025] The present invention enables objective and highly reliable risk assessment that does not rely on simple event detection by calculating a risk index based on quantitative measurements such as the rate of change in vehicle speed, duration of stoppage, rate of increase in smoke, and rate of increase in hazardous gas concentration. Accordingly, the present invention enables rapid automatic response at the initial warning stage of a tunnel accident.

[0026] The present invention distinguishes between the front and rear sections by reflecting tunnel section information and the vehicle's direction of travel, and selectively performs emergency broadcasting and communication control, thereby preventing unnecessary full-section warnings and minimizing driver confusion. The present invention also reduces the possibility of secondary accidents by calculating the expected smoke diffusion section considering the ventilation direction and expanding the response range.

[0027] The present invention ensures the continuity and reliability of emergency communication by establishing an independent transmission path for video of the accident section, automatically switching to a backup path in the event of communication delay, and activating adjacent section equipment as a replacement in the event of equipment failure. Accordingly, the present invention implements a stable and consistent automatic response system suitable for the special environment of tunnels. Brief explanation of the drawing

[0029] FIG. 1 is a block diagram illustrating the overall configuration of an emergency communication automatic interlocking system based on tunnel image recognition according to the present invention. FIG. 2 is a block diagram of an emergency response control terminal according to the present invention. FIG. 3 is a block diagram illustrating the detailed configuration of the risk analysis unit according to the present invention. FIG. 4 is a block diagram illustrating the detailed configuration of a section management unit according to the present invention. FIG. 5 is a block diagram illustrating the detailed configuration of a communication execution unit according to the present invention. FIG. 6 is a block diagram illustrating the detailed configuration of a network control unit according to the present invention. FIG. 7 is a block diagram illustrating the detailed configuration of an environment correction unit according to the present invention. FIG. 8 is a flowchart illustrating a tunnel image recognition-based emergency communication automatic interlocking method according to the present invention. Specific details for implementing the invention

[0030] Although the present invention may be subject to various modifications and may have various embodiments, specific embodiments are to be described in detail by way of illustration in the drawings. This is not intended to limit the present invention to specific embodiments, and it should be understood that any modification, equivalent, or substitution falls within the spirit and scope of the present invention for connecting and / or fixing structures extending in different directions.

[0031] The terms used in this specification are used merely to describe specific embodiments and are not intended to limit the invention. The singular expression includes the plural expression unless the context clearly indicates otherwise.

[0032] In this specification, terms such as “comprising” or “having” are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should not be understood as precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0033] Hereinafter, a preferred embodiment of the tunnel image recognition-based emergency communication automatic interlocking system according to the present invention will be described with reference to the accompanying drawings.

[0034] FIG. 1 is a block diagram illustrating the overall configuration of an emergency communication automatic interlocking system based on tunnel image recognition according to the present invention.

[0035] Referring to FIG. 1, the present invention includes a plurality of cameras (11) installed inside and outside a tunnel, an emergency terminal (12), a tunnel entrance control facility (20) that controls equipment installed at the tunnel entrance, and an emergency response control terminal (10) that collects and analyzes tunnel situation images in real time through a plurality of cameras (11) (Camera 1, Camera 2, ... Camera N) installed inside the tunnel.

[0036] Multiple cameras (11) are installed inside and / or outside the tunnel to capture images and transmit them to an emergency response control terminal (10).

[0037] The emergency terminal (12) is a communication terminal installed inside a tunnel, which automatically or manually establishes a call connection by a user or system when an emergency occurs. The emergency terminal (12) is connected to the emergency response control terminal (10) and automatically opens a call channel based on the control signal of the emergency response control terminal (10), and transmits status information (e.g., at least one of terminal identifier information, terminal installation location coordinate information, event type information, and danger level information) including terminal identification information and installation location information to a designated public institution. Accordingly, the external public institution system (30) can recognize the location and situation immediately upon receiving the call.

[0038] The tunnel entrance control equipment (20) controls barrier equipment (e.g., barrier bar, automatic barrier gate, lane blocking bollard) and variable information signs (e.g., electronic display board, LED warning light, lighting device), and can be linked with external traffic signals (e.g., traffic light turning red, detour guidance route display, traffic control center notification) and voice guidance equipment inside the tunnel (e.g., speaker), and can perform vehicle recognition-based selection control.

[0039] For example, the tunnel entrance control facility (20) blocks the entry of a vehicle into the tunnel by lowering the barrier bar or closing the automatic barrier gate based on an entry restriction control signal received from the entry interlocking module (540).

[0040] In addition, the tunnel entrance control equipment (20) displays a no-entry message on a variable information sign and lights up a warning light according to the entry restriction control signal.

[0041] Additionally, the tunnel entrance control equipment (20) transmits an entry restriction control signal to an external traffic signal control system to control the tunnel entry direction signal light to turn red or to display detour guidance route information.

[0042] In addition, the tunnel entrance control equipment (20) transmits an entry prohibition guidance voice through a speaker near the entrance in accordance with an entry restriction control signal.

[0043] Additionally, the tunnel entrance control equipment (20) receives vehicle identification information and, if it is determined to be an emergency vehicle, restricts the blocking operation and blocks entry only for general vehicles.

[0044] The emergency response control terminal (10) analyzes the video to determine whether there is a danger and controls the emergency terminal (12) installed inside the tunnel and the tunnel entrance control equipment (20).

[0045] Additionally, the emergency response control terminal (10) is connected via a communication network to an external public institution system (30) (e.g., a fire control system, a police control system, or a disaster safety integrated control system). At this time, the emergency response control terminal (10) transmits event type information, danger level information, and occurrence location coordinate information detected through the tunnel image to at least one of the external public institution systems (30).

[0046] Alternatively, the emergency response control terminal (10) receives reply information or dispatch confirmation information received from an external public institution system (30) and performs corresponding coordination (e.g., restriction of entry into tunnel, output of traffic light and electronic display information, activation of emergency terminal (12), switching of traffic light, display of detour route).

[0047] Additionally, the emergency response control terminal (10) is connected to an external traffic system (40) (e.g., an external traffic signal control system, a road traffic control system, a variable information sign control system) and transmits an entry restriction control signal to the external traffic signal control system when the situation inside the tunnel satisfies the set entry restriction conditions through video.

[0048] At this time, the external traffic signal control system receives entry restriction information from the emergency response control terminal (10) and switches the tunnel entry direction signal light to red, the variable information sign control system displays a detour guidance message on the electronic display board at the tunnel entrance, and the road traffic control system performs a detour route display for the upper road network.

[0049] The emergency response control terminal (10) is described with reference to FIG. 2.

[0050] FIG. 2 is a block diagram of an emergency response control terminal according to the present invention.

[0051] Referring to FIG. 2, the emergency response control terminal (10) analyzes video and various measurement data collected within the tunnel to determine the risk, and automatically controls emergency communication equipment and field equipment according to the determination result. To this end, the emergency response control terminal (10) includes a risk analysis unit (100), a section management unit (200), a communication execution unit (300), a network control unit (400), and an environment correction unit (500).

[0052] The risk analysis unit (100) analyzes video data received from the camera (11) to track vehicle objects and recognize events, and calculates a quantitative risk index. When multiple events occur in the same section, the risk analysis unit (100) transmits the calculated risk index, which is calculated by accumulating the weights of each event, to the section management unit (200).

[0053] The section management unit (200) divides the tunnel into multiple sections and assigns an identifier to each section to map the location where an event occurs to the corresponding section, determines the direction of vehicle movement, and distinguishes between the front section and the rear section.

[0054] Here, the section management unit (200) compares the risk index and the reference value, sets the section to be responded to, and expands the response range to adjacent sections by reflecting ventilation direction information. The section management unit (200) transmits the set response section information to the communication execution unit (300).

[0055] The response target section refers to a control target area within the tunnel established based on the location where a dangerous event occurred, and the said response target section is dynamically set based on at least one of event occurrence location information, danger level information, or diffusion range information calculated by the environment correction unit. That is, the response target section is used as a reference area for controlling broadcasting equipment, activating emergency terminals, setting notification ranges, or determining entry restrictions.

[0056] The communication execution unit (300) activates the emergency terminal (12) for the set response section and transmits an emergency broadcast to the selected section. The communication execution unit (300) opens an emergency call when necessary and transmits danger information to the external traffic system (40) and the control center. The communication execution unit (300) selects a response scenario according to the danger level and gradually increases the response level.

[0057] The network control unit (400) applies priority transmission processing to the video data of the accident section. The network control unit (400) limits the bandwidth of general data traffic and secures an emergency communication band, and switches to a backup path if the communication delay exceeds a threshold value. That is, the network control unit (400) monitors the communication status and resets the transmission path when an error occurs.

[0058] The environment correction unit (500) calculates the expected smoke diffusion section by reflecting the operating direction of the ventilation equipment, expands the response range for the expected diffusion section, and corrects the broadcast sound by reflecting the reverberation characteristics of the tunnel. Here, the environment correction unit (500) applies a weight to the risk index by reflecting the characteristics of the enclosed space.

[0059] That is, the emergency response control terminal (10) calculates a risk index after video collection, sets a response section according to the risk index, executes emergency communication in the selected section, ensures communication stability, and adjusts the response range by reflecting environmental conditions, thereby proceeding with the entire process.

[0060] Below, a more detailed description of each component of the emergency response control terminal described above is provided.

[0061] FIG. 3 is a block diagram illustrating the detailed configuration of the risk analysis unit according to the present invention.

[0062] Referring to FIG. 3, the risk analysis unit (100) includes an image acquisition module (110), an object tracking module (120), an event recognition module (130), and a risk calculation module (140).

[0063] The image acquisition module (110) receives image data from a plurality of cameras (11). The image acquisition module (110) divides the image data into frames and combines camera (11) identification information and shooting location information with each frame. The image acquisition module (110) performs time alignment of frames by aligning time information for each camera (11).

[0064] The image acquisition module (110) calculates an average brightness value, a contrast variance value, a blur index, and a noise ratio for each frame. The image acquisition module (110) compares the calculated image quality index with a reference value. If the average brightness value is outside the reference range, the blur index is greater than or equal to the reference value, or the noise ratio is greater than or equal to the reference value, the image acquisition module (110) sets a quality degradation flag for the corresponding frame.

[0065] The image acquisition module (110) stores a reference frame and defines a fixed area by comparing it with the current frame. The image acquisition module (110) calculates the pixel change rate within the fixed area and determines the light source change state if the change rate is greater than or equal to the reference rate.

[0066] The object tracking module (120) detects a vehicle object in the video data input from the video acquisition module (110) and calculates the position coordinates of the vehicle object. The object tracking module (120) generates a movement trajectory using the position change of the vehicle object between frames and calculates velocity, acceleration, and direction of travel vectors based on the movement trajectory. The object tracking module (120) determines whether the vehicle object deviates from the lane by referring to lane area information and assigns a weight that increases the possibility of driving in the wrong direction if lane deviation occurs continuously.

[0067] The event recognition module (130) determines whether an event has occurred by using movement information transmitted from the object tracking module (120) and image feature information transmitted from the image collection module (110). For example, the event recognition module (130) determines that it is a vehicle stop event if the vehicle speed is below a reference speed and the state persists for a reference time or longer. Alternatively, the event recognition module (130) determines that it is a reverse driving event if the direction of travel vector is opposite to the reference direction and the state persists for a reference number of frames or longer.

[0068] Additionally, the event recognition module (130) divides the smoke candidate region within the image and calculates the area increase rate and concentration change rate of the smoke candidate region, and calculates the brightness change amount and color distribution change of the flame candidate region. At this time, if the smoke candidate region or the flame candidate region overlaps with a fixed region by more than a certain ratio, the event recognition module (130) excludes the region from the candidate region. Here, the event recognition module (130) classifies events where the number of consecutive frames is less than a reference value as temporary noise.

[0069] Additionally, the event recognition module (130) calculates an event reliability value by referring to a quality degradation flag transmitted from the image acquisition module (110). The event recognition module (130) calculates the reliability value by normalizing at least one of the frame continuity index, the object tracking consistency index, and the area change stability index.

[0070] The event recognition module (130) generates a composite event flag when multiple events occur simultaneously in the same section and transmits the event type, time of occurrence, duration, occurrence coordinates, and reliability value to the risk calculation module (140).

[0071] The risk calculation module (140) receives event information and reliability values ​​transmitted from the event recognition module (130). Additionally, the risk calculation module (140) uses at least one of the vehicle speed reduction rate, stop duration, smoke area increase rate, and flame brightness change rate as a risk calculation factor. The risk calculation module (140) calculates a risk index by applying weights for each event type, normalizing each factor, and then performing a weighted summation. Here, the risk calculation module (140) reflects the reliability value in the weights and, if a composite event flag is set, applies the weights of each event cumulatively.

[0072] In addition, the risk calculation module (140) classifies the calculated risk index into multiple stages.

[0073] For example, the risk calculation module (140) sets the risk index to a warning stage when the risk index is greater than or equal to a first threshold value or less than a second threshold value, and sets the risk index to an emergency stage when the risk index is greater than or equal to the second threshold value. If the risk index is maintained below the first threshold value for a certain period of time, it returns to a normal stage.

[0074] The risk calculation module (140) transmits at least one of the event type, time of event occurrence, frame number of event occurrence, event duration, coordinates of event occurrence, section identifier, direction of vehicle travel, reliability value, risk index value, and risk level information to the section management unit (200).

[0075] In this way, the risk analysis unit (100) performs image quality judgment, object tracking, event recognition, reliability calculation, false detection suppression, complex event accumulation, staged risk judgment, and segment matching information transmission in stages. This structure is distinguished from a simple image detection method and provides a quantitative-based risk judgment structure.

[0076] FIG. 4 is a block diagram illustrating the detailed configuration of a section management unit according to the present invention.

[0077] Referring to FIG. 4, the section management unit (200) includes a structure storage module (210), a section division module (220), a direction analysis module (230), and a range determination module (240).

[0078] The structure storage module (210) stores physical structure information of the tunnel. The structure storage module (210) includes information on the total length of the tunnel, the number of lanes, the location of entrances and exits, the location of ventilation equipment, the location of emergency terminals (12), and the location of camera installation (11).

[0079] For example, if the tunnel length is 1,200m and three cameras (11) each photograph sections of 0~400m, 350~800m, and 750~1,200m, the structure storage module (210) stores the shooting range and overlapping area information of each camera (11).

[0080] Additionally, the structure storage module (210) stores conversion coefficients between the camera (11) image coordinate system and the tunnel actual distance coordinate system. For example, if 1 pixel of the image coordinates' x-axis corresponds to an actual distance of 0.05 m, the corresponding conversion coefficient is recorded.

[0081] The section division module (220) divides the tunnel into multiple management sections.

[0082] For example, a tunnel length of 1,200m is divided into 100m units to create a total of 12 sections. Each section is assigned a section identifier (G1~G12).

[0083] As another example, when dividing the sections based on the shooting range of the camera (11), the sections are set as 0~350m, 350~750m, 750~1,200m based on the overlapping area of ​​the camera (11).

[0084] Additionally, the section division module (220) subdivides the section when the traffic volume increases above a reference value in a specific section. For example, if the vehicle density of section G5 is 1.5 times or more than the average, G5 is divided into G5-1 and G5-2.

[0085] The direction analysis module (230) receives the vehicle's direction of travel vector and compares it with the tunnel's reference direction.

[0086] For example, if the tunnel standard direction of travel is from the entrance to the exit (+) and the vehicle travel vector is maintained in the (-) direction for a certain period of time or longer, it is determined to be in the reverse direction. Specifically, the direction analysis module (230) sets the forward sections (G7, G8) as priority response targets if the point of danger is section G6 and the vehicle travel direction is forward. Conversely, if it is in the reverse direction, sections G5 and G4 are set as priority response targets.

[0087] The range determination module (240) receives risk index and risk level information transmitted from the risk calculation module (140).

[0088] For example, the range determination module (240) classifies the risk index as 60, the first threshold value as 50, and the second threshold value as 80, thereby classifying it as a warning stage. And when the range determination module (240) is in the warning stage, it includes the occurrence section (G6) and adjacent sections (G5, G7).

[0089] Alternatively, the range determination module (240) classifies the situation as an emergency stage when the risk index is 85. In addition, the range determination module (240) includes the occurrence section (G6), adjacent sections (G5, G7), and two preceding sections (G8, G9) when the situation is an emergency stage.

[0090] Alternatively, the range determination module (240) calculates the estimated smoke diffusion time when the ventilation direction is the exit direction. For example, if the smoke diffusion speed is 1 m / s and the estimated diffusion time is 120 seconds, an additional 120 m range is included. Accordingly, one or two additional sections after G6 are included.

[0091] Thus, the range determination module (240) generates a list of final corresponding target sections and transmits a list of facility identifiers of the corresponding sections to the communication execution unit (300).

[0092] In this way, the section management unit (200) performs coordinate transformation-based position matching, distance-based section division, traffic volume-based dynamic segmentation, direction vector-based forward / rear distinction, and risk stage and diffusion time-based range expansion. This structure provides a quantitative section management structure that reflects tunnel structure information and time-based diffusion prediction.

[0093] FIG. 5 is a block diagram illustrating the detailed configuration of a communication execution unit according to the present invention.

[0094] Referring to FIG. 5, the communication execution unit (300) receives response target section information transmitted from the section management unit (200) and risk level information transmitted from the risk calculation module (140), and sequentially performs response procedures according to the input information.

[0095] To this end, the communication execution unit (300) includes a scenario module (310), a broadcast control module (320), a telephone control module (330), a notification control module (340), and a stage control module (350).

[0096] The scenario module (310) determines a control policy within the communication execution unit (300) and generates a control data packet.

[0097] For example, the scenario module (310) receives event type information from the input event recognition module (130), risk level information and risk index value from the risk calculation module (140), and a list of corresponding target range identifiers from the range determination module (240), and queries a response matrix based on the combination of event type and risk level. Then, the scenario module (310) determines the control items and execution order defined in the response matrix (for example, if the vehicle stop event is in the warning level, only the deceleration guidance broadcast is performed).

[0098] Alternatively, the scenario module (310) determines to simultaneously perform evacuation broadcasting, automatic connection to emergency terminal (12), and notification to external agencies when the fire event is in an emergency phase.

[0099] Alternatively, the scenario module (310) generates a broadcast control data packet when broadcast control is selected. Here, the broadcast control data packet includes at least one of a packet identifier, an event type code, a risk level code, a list of corresponding target segment identifiers, a transmission message code, a voice data size, and generation time information.

[0100] The scenario module (310) transmits the generated broadcast control data packet to the stage control module (350), and the stage control module (350) transmits the packet to the broadcast control module (320).

[0101] Additionally, if the response policy includes a phone connection item, the scenario module (310) generates a phone control data packet by querying the emergency terminal (12) identifier mapped to each segment stored in the structure storage module (210) based on the list of response target segment identifiers. Then, the scenario module (310) transmits the generated phone control data packet to the stage control module (350). The stage control module (350) transmits the packet to the phone control module (330).

[0102] Here, the phone control data packet includes at least one of a packet identifier, a segment identifier, an emergency terminal (12) identifier, a control center connection number, a connection mode code, and a session duration.

[0103] Additionally, the scenario module (310) generates a notification data packet when the response policy includes an external notification item. The notification data packet includes at least one of event type information, risk level information, risk index value, event occurrence coordinate information, and a list of response target range identifiers. The scenario module (310) transmits the generated notification data packet to the notification control module (340).

[0104] The broadcast control module (320) receives from the section management unit (200) one of the following: a list of corresponding target section identifiers, start and end coordinates of each section, a list of speaker identifiers corresponding to each section, risk level information, and event type information.

[0105] The list of corresponding target range identifiers is generated by the range determination module (240) of the range management unit (200). Here, the range determination module (240) receives risk index and risk level information transmitted from the risk calculation module (140), generates the final list of corresponding target ranges, and transmits it to the communication execution unit (300).

[0106] Additionally, the start and end coordinates of each section are generated by the section division module (220). Here, the section division module (220) sets the start and end coordinates of each section based on tunnel length information and stores them in a database. Then, when the range determination module (240) generates a list of corresponding target section identifiers, it queries the start and end coordinates of the corresponding sections together and transmits them to the communication execution unit (300).

[0107] Speaker identifier information corresponding to each section is generated in the structure storage module and transmitted to the communication execution unit (300) through the range determination module (240). Here, the structure storage module (210) assigns a unique identifier to each piece of equipment (speaker, emergency terminal (12), etc.) and stores the equipment location coordinates, and generates a section-equipment mapping table by comparing the equipment location coordinates with the section coordinates. Then, after the corresponding target section is determined, the range determination module (240) retrieves the list of equipment identifiers mapped to the corresponding section and transmits it to the communication execution unit (300).

[0108] Risk level information is generated by the risk calculation module (140) and transmitted to the communication execution unit (300). Here, the risk calculation module (140) calculates a risk index using the event information and reliability value transmitted from the event recognition module (130) and determines the risk level by comparing it with a reference value. Then, the risk calculation module (140) transmits the determined risk level information to the interval management unit (200), and the range determination module (240) transmits the risk level information, for which a response range has been set using the received risk level information, to the communication execution unit (300).

[0109] Event type information is generated by the event recognition module (130) and transmitted to the communication execution unit (300) by the range determination module (240). Here, the event recognition module (130) determines the corresponding event type among vehicle stop, reverse driving, smoke, and flame, and transmits it to the risk calculation module (140). Therefore, the risk calculation module (140) transmits the event type information to the section management unit (200) along with the determination of the risk level.

[0110] Then, the broadcast control module (320) queries the speaker mapping table stored in the database based on the input section identifier list, and determines the transmission message using the retrieved speaker identifier and risk level information.

[0111] For example, the broadcast control module (320) selects a deceleration guidance message when the danger level is the warning level and the event type is vehicle stop. The broadcast control module (320) selects an evacuation guidance message when the danger level is the emergency level and the event type is fire.

[0112] Additionally, the broadcast control module (320) converts the selected message into voice data. At this time, the broadcast control module (320) activates the output channel corresponding to each speaker identifier. Then, the broadcast control module (320) transmits the voice data through the activated output channel.

[0113] Additionally, the broadcast control module (320) prevents duplicate transmission by integrating speaker identifiers when adjacent sections overlap. The broadcast control module (320) records whether the transmission was successful after the transmission is completed.

[0114] The phone control module (330) receives a phone control data packet from the scenario module (310) and activates the emergency terminal (12). At this time, the phone control module (330) establishes a call session of the emergency terminal (12), performs a reconnection in the event of a connection failure, and transmits the execution result to the stage control module (350).

[0115] Specifically, the telephone control module (330) checks the network address of the emergency terminal (12) located in the corresponding section by querying the emergency terminal (12) mapping table stored in the structure storage module (210) based on the section identifier received through the telephone control data packet. Then, the telephone control module (330) transmits a connection readiness signal to the emergency terminal (12) and checks whether the emergency terminal (12) returns a response signal. For example, the telephone control module (330) activates the emergency terminal (12) TEL-06 corresponding to section G6 and automatically dials the control center number.

[0116] At this time, the telephone control module (330) creates a session between the emergency terminal (12) and the control center, starts a session timer, and terminates the session when the session maintenance time elapses or a call termination signal is received.

[0117] Additionally, the phone control module (330) records the number of failures when a connection failure occurs, and performs a reconnection if the number of failures is less than a threshold value. If the number of failures is greater than or equal to the threshold value, the phone control module (330) transmits the failure information to the stage control module (350). If the failure information exceeds the threshold value, the stage control module (350) transmits a re-evaluation request signal to the risk calculation module (140).

[0118] Additionally, the phone control module (330) transmits whether the call connection was successful to the step control module (350). Upon receiving this, the step control module (350) records whether the connection was successful and updates the execution status.

[0119] The notification control module (340) receives a notification data packet generated by the scenario module (310) and received through the stage control module (350). Based on the received notification data packet (at least one of event type, risk index, risk stage, occurrence coordinates, estimated spread distance, and list of response target sections), the notification control module (340) constructs a transmission message that conforms to the specifications of an external public institution system (30) and / or an external traffic system (40). At this time, the notification control module (340) converts the coordinate information into a latitude and longitude or tunnel distance standard format. For example, the notification control module (340) converts a tunnel distance of 620m into GPS coordinates and includes them in the transmission message.

[0120] The notification control module (340) transmits the generated message to an external traffic system (40) or an external public institution system and receives a response code. At this time, if the response code is not a normal value, the notification control module (340) performs retransmission. Then, if the number of retransmissions exceeds a threshold value, the notification control module (340) transmits failure information to the stage control module (350). When failures accumulate, the stage control module (350) transmits a re-evaluation request signal to the risk calculation module (140).

[0121] Additionally, the notification control module (340) transmits a deceleration or entry blocking signal to the forward section traffic control device, including information on the expected spread distance. For example, the notification control module (340) requests an external system to restrict entry signals for sections G8 and G9 when a fire occurs in G6.

[0122] The stage control module (350) coordinates the timing and execution order of control execution within the communication execution unit (300). For example, the scenario module (310) transmits a generated packet among broadcast control data packets, telephone control data packets, and external notification data packets to the stage control module (350). The stage control module (350) analyzes the event type code and risk stage code of each data packet transmitted from the scenario module (310).

[0123] And the stage control module (350) sets the broadcast control data packet as the highest priority execution target when the danger stage information is the emergency stage. At this time, when the danger stage information is the warning stage, the stage control module (350) sets only the broadcast control data packet as the execution target and suspends telephone control and external notification.

[0124] For example, the stage control module (350) sets the execution targets in the order of broadcast control data packet, telephone control data packet, and external notification data packet when the fire event is in the emergency stage. Then, the stage control module (350) registers the execution order in the execution waiting queue.

[0125] Additionally, the stage control module (350) transmits execution signals to each control module according to the execution order. In the case of an emergency stage, the stage control module (350) executes telephone control after a set time interval (e.g., 3 seconds) has elapsed following the execution of the broadcast. Alternatively, in the case of a warning stage, the stage control module (350) sets a long interval for repeated broadcasts. The stage control module (350) records each execution time.

[0126] Additionally, the stage control module (350) generates a network control request data packet by calculating the data size, estimated transmission time, and allowable delay time of the data packet scheduled for execution. The network control request data packet includes at least one field among a data type code, a required bandwidth value, a priority level, a maximum allowable delay time, and an estimated transmission duration time. For example, if the emergency broadcast data is 500MB and must be transmitted within 30 seconds, the stage control module (350) calculates the required bandwidth as 133Mbps. Then, the stage control module (350) transmits the generated network control request data packet to the network control unit (400).

[0127] Additionally, the stage control module (350) can process feedback of the execution results. For example, the broadcast control module (320) transmits to the stage control module (350) whether transmission is complete, and the telephone control module (330) transmits to the stage control module (350) whether the call connection is successful. The notification control module (340) transmits to the stage control module (350) whether transmission is successful. Here, the stage control module (350) determines whether to re-execute if an execution failure occurs, and if the number of re-executions exceeds a threshold value, transmits a signal requesting re-evaluation to a higher risk level to the risk calculation module (140).

[0128] At this time, the risk calculation module (140) classifies a serious communication anomaly as a broadcast transmission failure that occurs more than a threshold value (e.g., 3 times or more) and adjusts the risk weight. At this time, the risk weight adjustment proceeds in one of the following ways: increasing the weight in proportion to the number of control failures, adding a fixed weight when a specific facility control fails, or increasing the cumulative weight when the same section fails repeatedly.

[0129] For example, the risk calculation module (140) adds a weight of 10 to the existing risk index if an external notification failure occurs during a fire event. Alternatively, the risk calculation module (140) adds a weight of 20% to the existing risk index if there are 5 or more broadcast failures.

[0130] Alternatively, the risk calculation module (140) recalculates a new risk index by reflecting the adjusted weights to the existing risk index value. The risk calculation module (140) applies the following equation.

[0131] Recalculated Risk Index = Original Risk Index + (Number of Failures × Failure Weighting Factor)

[0132] For example, if the existing risk index is 75, the failure weighting coefficient is 3, and the number of failures is 4, the risk calculation module (140) calculates a new risk index of 87.

[0133] Then, the risk calculation module (140) compares the recalculated risk index with a reference value, and if the recalculated risk index is greater than or equal to the second reference value, it raises the risk level to the emergency level. Alternatively, if the recalculated risk index is greater than or equal to the first reference value but less than the second reference value, the risk calculation module (140) maintains it at the warning level. Then, if a level change occurs, the risk calculation module (140) transmits a level change signal to the section management unit (200) and the level control module (350). Then, the section management unit (200) receives the level change signal and expands the corresponding target section.

[0134] Additionally, the stage control module (350) monitors changes in the stage. For example, the risk calculation module (140) transmits a change signal to the stage control module (350) when the risk stage rises or falls. The stage control module (350) reduces the delay time between executions when the risk stage rises.

[0135] Alternatively, the stage control module (350) stops some control or extends the repetition interval when the risk level is lowered. The stage control module (350) records the time of the stage change and the history of the control change.

[0136] FIG. 6 is a block diagram illustrating the detailed configuration of a network control unit according to the present invention.

[0137] Referring to FIG. 6, the network control unit (400) receives corresponding target section information, risk level information, and control execution status information transmitted from the communication execution unit (300), and redistributes communication network resources (e.g., at least one of transmission bandwidth, transmission queue buffer capacity, packet priority label, optical path channel number, VLAN identifier, wireless backhaul channel band, number of network equipment ports allocated, number of simultaneous transmission sessions) based on the received information.

[0138] To this end, the network control unit (400) includes a priority control module (410), a band control module (420), a path switching module (430), a status monitoring module (440), and an emergency channel creation module (450).

[0139] The priority control module (410) receives a network control request data packet from the stage control module (350) that includes at least one of a data type code, required bandwidth, priority level, and maximum allowable delay time, compares the risk stage information with the priority level, and if the priority level is greater than or equal to a reference value, places the data packet in the highest priority transmission queue. For example, the priority control module (410) sets a fire emergency stage packet to priority level 7 and registers it in the top queue. The priority control module (410) moves general data packets with lower priority to a lower queue.

[0140] That is, the first control module (410) places the emergency control data packet in the upper queue among the multiple transmission queues and moves the general data packet to the lower queue based on the priority level information included in the network control request data packet received from the stage control module (350), thereby resetting the transmission priority.

[0141] The band control module (420) is characterized by being configured to reallocate network bandwidth and transmission queue resources to minimize transmission delay of emergency control data by comparing a network control request data packet containing a requested bandwidth value received from the stage control module (350) with current bandwidth usage and port occupancy information received from the status monitoring module (440).

[0142] For example, the stage control module (350) calculates a minimum required bandwidth of 133 Mbps to transmit 500 MB of broadcast data within 30 seconds for a fire event in the emergency stage and sets the priority level to 7. Meanwhile, the status monitoring module (440) collects information on current bandwidth usage, port occupancy, and queue occupancy from network equipment. Then, the status monitoring module (440) generates current network resource status information based on the collected information and transmits it to the band control module (420). For example, the status monitoring module (440) calculates that 900 Mbps of the total 1 Gbps bandwidth is in use and that a specific port occupancy is 85%, and transmits this to the band control module (420).

[0143] The band control module (420) refers to the required bandwidth value and priority level information included in the network control request data packet, and simultaneously analyzes the network resource status information received from the status monitoring module (440). Then, when the priority level is greater than or equal to the reference value, the band control module (420) reallocates network resources by limiting the transmission rate of general traffic, redistributing transmission queue capacity, or resetting VLAN resources in order to minimize the transmission delay of emergency control data (for example, allocating 300 Mbps for emergency control data, limiting the general data transmission rate to 600 Mbps, creating VLAN 200 dedicated to emergency data, and assigning priority bandwidth to the VLAN).

[0144] The path switching module (430) receives information on the average delay time and packet loss rate of the main transmission path from the status monitoring module (440). The phase control module (350) includes information on the maximum allowable delay time in a network control request data packet and transmits it to the network control unit (400) in order to transmit emergency broadcasts, telephone connections, or external notification data within a set time.

[0145] Then, the path switching module (430) compares the maximum allowable delay time with the average delay time of the main transmission path and checks the status of the backup transmission path if the average delay time exceeds the maximum allowable delay time or if the packet loss rate exceeds the set reference loss rate.

[0146] For example, if the status monitoring module (440) measures the average delay time of the main optical path λ1 as 150ms and the stage control module (350) sets the maximum allowable delay time for emergency broadcast data to 50ms, the path switching module (430) determines that the main path has exceeded the allowable criteria. Then, if the average delay time of the reserve optical path λ2 is 40ms and the packet loss rate is less than 1%, the path switching module (430) switches the transmission session to the reserve path. Afterward, the path switching module (430) releases the existing path after receiving a switching completion signal and transmits the switching result to the stage control module (350).

[0147] That is, the path switching module (430) is characterized by being configured to switch the transmission session to a backup transmission path if the delay time or loss rate exceeds the set criteria by comparing the delay time and packet loss rate information of the main transmission path received from the state monitoring module (440) with the maximum allowable delay time information received from the stage control module (350).

[0148] The status monitoring module (440) collects information on transmission bandwidth usage, latency, packet loss rate, and port occupancy from network equipment, generates network resource status information based on the information, and transmits it to the priority control module (410), the band control module (420), and the path switching module (430).

[0149] The status monitoring module (440) periodically receives real-time transmission bandwidth usage, average delay time, packet loss rate, and port occupancy information from the network switch and optical transmission equipment.

[0150] For example, the status monitoring module (440) receives that 900 Mbps of the total 1 Gbps bandwidth is in use, the average delay time of the main optical path λ1 is 120 ms, the packet loss rate is 4%, and the specific port occupancy rate is 85%.

[0151] The state monitoring module (440) calculates the average value of the received data over a reference time interval (e.g., 5 seconds). Then, the state monitoring module (440) compares the calculated average value with a preset threshold value. For example, the state monitoring module (440) determines an abnormal state if the average delay time exceeds 100ms or the packet loss rate exceeds 3%.

[0152] The status monitoring module (440) generates network resource status information including a judgment result. The status monitoring module (440) then transmits the generated network resource status information to the priority control module (410) to be used for determining queue reallocation, transmits it to the band control module (420) to be used for determining bandwidth reallocation, and transmits it to the path switching module (430) to be used for determining whether to switch to a preliminary path. The status monitoring module (440) periodically updates the network resource status information and repeatedly transmits it to each module.

[0153] The emergency channel creation module (450) creates a dedicated virtual transmission channel separated from existing communication traffic according to the emergency channel creation request signal received from the stage control module (350) and sets it to transmit emergency control data packets through the dedicated channel.

[0154] For example, the stage control module (350) generates an emergency channel creation request signal when the danger stage is determined to be an emergency stage, and transmits the request signal to the network control unit (400) including the required bandwidth value and priority level information.

[0155] Then, the emergency channel creation module (450) receives a request signal. Then, the emergency channel creation module (450) queries the VLAN configuration information of the current network and the optical path channel usage status.

[0156] For example, the emergency channel creation module (450) creates VLAN 200 exclusively for emergency control data when existing general traffic uses VLAN 100. Additionally, the emergency channel creation module (450) allocates a new wavelength channel λ3 in the optical transmission equipment and configures emergency broadcast and notification data packets to be transmitted only through VLAN 200 and channel λ3. Then, the emergency channel creation module (450) transmits a channel creation completion signal to the stage control module (350).

[0157] The prediction resource reservation module (460) receives the estimated distance and estimated time of spread when a fire event occurs calculated by the risk calculation module (140), and calculates the possibility of additional broadcast transmission or external notification occurring during the estimated time of spread, and the estimated amount of data transmission during the estimated time of spread.

[0158] For example, if the risk calculation module (140) calculates the expected spread time as 120 seconds and two additional broadcast transmissions are expected, with each transmission data being 200MB, the predicted resource reservation module (460) preemptively reserves a bandwidth of approximately 27Mbps or more for 120 seconds for a total transmission of 400MB. Then, the predicted resource reservation module (460) separates the reserved bandwidth from general traffic and sets it as a resource dedicated to emergency control data.

[0159] The independent verification module (470) compares the primary communication network status information received from the status monitoring module (440) with the secondary communication network status information remeasured through a separate path, and determines the communication network abnormal state if the difference between the two information exceeds a set standard.

[0160] For example, the independent verification module (470) receives information on the average delay time and packet loss rate of the main transmission path generated by the state monitoring module (440), and transmits multiple test packets for the same transmission path to measure the secondary delay time. Here, the independent verification module (470) collects the delay time values ​​measured multiple times, calculates the average value of the delay time values, calculates the standard deviation, and removes values ​​exceeding the set deviation standard relative to the average value as outliers to calculate the secondary average delay time.

[0161] For example, if the independent verification module (470) performs 5 measurements on the same path and obtains delay times of 58ms, 61ms, 60ms, 59ms, and 310ms, the independent verification module (470) removes 310ms as an outlier and calculates a secondary average delay time of about 59.5ms for the remaining values.

[0162] At this time, the independent verification module (470) determines the second average delay time as a valid measurement value when the standard deviation of the second average delay time is below the set standard and compares the first average delay time with the second average delay time. Then, if the difference between the two values ​​exceeds the set standard and the second average delay time falls within the normal range, the independent verification module (470) generates a first measurement error flag and transmits it to the band control module (420) and transmits a route switching hold signal to the route switching module (430). Here, when the band control module (420) receives the first measurement error flag, it suspends network resource reallocation and maintains the existing band structure.

[0163] Conversely, if both the first average delay time and the second average delay time exceed the maximum allowable delay time criteria, the independent verification module (470) generates a network anomaly confirmation signal and transmits it to the path switching module (430), and at the same time transmits the network reliability degradation index to the risk calculation module (140).

[0164] The path switching module (430) switches the transmission session to a backup transmission path when it receives a network abnormality confirmation signal.

[0165] The risk calculation module (140) calculates a recalculated risk index by reflecting the network reliability degradation index as a weight to the existing risk index. For example, if the existing risk index is 75 and the network reliability degradation index is 10, the risk calculation module (140) calculates the recalculated risk index as 85 and raises the risk level.

[0166] In this way, the independent verification module (470) determines the reliability of secondary measurements through multiple actual measurement-based statistical processing and outlier removal, and reflects this in the band control, path switching, and risk index calculation processes, thereby forming a closed-loop control structure that integrates network control and risk judgment.

[0167] The section weighting adjustment module (480) receives section-specific accident history information and traffic density information, calculates section-specific weighting coefficients, and reflects the weighting coefficients in the priority level and bandwidth allocation ratio of the emergency control data packet.

[0168] To explain in more detail, the section weighting adjustment module (480) receives information on the number of accidents per section and average traffic density information stored in the database and calculates a risk contribution coefficient for each section. Then, the section weighting adjustment module (480) compares the accident frequency and average traffic volume over a recent period with reference values, and calculates a section weighting coefficient by reflecting the accident frequency increase rate and traffic density ratio.

[0169] For example, if five accidents occurred in section G6 in the last year and the average traffic density is 1.4 times the reference value, the section weighting adjustment module (480) calculates a final weighting coefficient of 1.3 by reflecting the accident frequency coefficient of 1.2 and the traffic density coefficient of 1.1.

[0170] The section weight adjustment module (480) transmits the calculated weight coefficients to the priority control module (410).

[0171] Then, the first control module (410) increases the priority level of the emergency control data packet generated in the G6 section by reflecting the weighting coefficient. The section weighting adjustment module (480) transmits the calculated weighting coefficient to the band control module (420).

[0172] The band control module (420) applies a higher bandwidth allocation ratio to the emergency control data in the G6 section by reflecting a weighting factor. For example, when the basic allocation ratio is 20%, the band control module (420) applies a weighting factor of 1.3 to the G6 section to prioritize the allocation of approximately 26% of the bandwidth.

[0173] The section weight adjustment module (480) compares the weight coefficients of each section when the same event occurs in multiple sections and transmits priority information to the stage control module (350) to designate the section with a relatively higher weight coefficient as the priority corresponding section.

[0174] Additionally, the interval weight adjustment module (480) transmits the interval risk accumulation index to the risk calculation module (140) when an increase in the repetitive accident history is detected.

[0175] Then, the risk calculation module (140) calculates a recalculated risk index by reflecting the interval risk cumulative index as a weight to the existing risk index. For example, if the existing risk index is 70 and the interval risk cumulative index is 8, the risk calculation module (140) calculates the recalculated risk index as 78.

[0176] In this way, the section weight adjustment module (480) forms a closed-loop control structure that differentiates communication control and risk analysis according to section characteristics by being involved in calculating section history-based weights, correcting priorities, adjusting bandwidth ratios, and correcting risk indices.

[0177] FIG. 7 is a block diagram illustrating the detailed configuration of an environment correction unit according to the present invention.

[0178] Referring to FIG. 7, the environment correction unit (500) includes a diffusion prediction module (510), an acoustic correction module (520), a risk correction module (530), and an entry linkage module (540).

[0179] The environment correction unit (500) includes a diffusion prediction module (510), an acoustic correction module (520), a risk correction module (530), and an entry linkage module (540).

[0180] For example, the diffusion prediction module (510) receives fire event occurrence location information and event intensity information calculated by the risk calculation module (140), and calculates the risk diffusion path by combining the tunnel internal wind speed and airflow direction information received from the environment sensor.

[0181] The diffusion prediction module (510) calculates the smoke diffusion distance by applying a tunnel length coordinate system and an airflow direction vector. For example, if the diffusion prediction module (510) determines that the fire origin point is at the 600m point of the tunnel, the wind speed is 3m / s, and the airflow is formed in the downward direction, the diffusion prediction module (510) calculates the smoke diffusion distance after 120 seconds to be approximately 360m.

[0182] The diffusion prediction module (510) transmits the calculated diffusion distance and diffusion time to the range determination module (240) and the prediction resource reservation module (460). At this time, the range determination module (240) receives the estimated diffusion distance and estimated diffusion time received from the diffusion prediction module (510), and calculates a plurality of section identifiers included in the estimated diffusion range by referring to the tunnel section length information stored in the structure storage module (210) and the section boundary information defined by the section division module (220). That is, the range determination module (240) determines and / or readjusts the corresponding target section according to the diffusion distance and diffusion time of the diffusion prediction module (510).

[0183] Additionally, the prediction resource reservation module (460) calculates and / or recalculates the number of communication facilities, the number of control operations, and the amount of transmission data in the section included in the spreading range based on the spreading time and spreading distance information received from the spreading prediction module (510).

[0184] The acoustic correction module (520) receives tunnel cross-section structure and reflection coefficient information from the structure storage module (210) and receives current background noise level information from the environment sensor to adjust the output intensity of the emergency broadcast.

[0185] Here, the acoustic correction module (520) distinguishes between sections with high and low tunnel reflectance and calculates a sound pressure level suitable for each section. For example, if the acoustic correction module (520) measures background noise at 85 dB, it sets the sound pressure of the emergency announcement broadcast to 95 dB, and limits the output amplification rate to 1.1 times in sections with high reflectance. The acoustic correction module (520) transmits the adjusted sound pressure setting value to the broadcast control module (320).

[0186] The risk correction module (530) receives the estimated diffusion distance and estimated diffusion time from the diffusion prediction module (510), and calculates a risk index correction coefficient by reflecting the traffic density and ventilation status information inside the tunnel.

[0187] At this time, the risk correction module (530) decreases the correction factor when the ventilation device is operating normally and increases the correction factor when the ventilation device is stopped. For example, if the risk calculation module (140) calculates the basic risk index as 70 and a ventilation stop state is detected, the risk correction module (530) applies a correction factor of 1.2 to recalculate the corrected risk index as 84. Then, the risk correction module (530) transmits the corrected risk index to the stage control module (350).

[0188] The entry linkage module (540) determines whether to restrict entry to the tunnel by combining the expected spread distance received from the spread prediction module (510) and the corrected risk index received from the risk correction module (530).

[0189] At this time, the entry interlock module (540) generates a tunnel entrance blocking signal when the correction risk index exceeds the set standard. For example, if the expected spread distance is 400m and the correction risk index is calculated to be 85, the entry interlock module (540) closes the tunnel entrance blocking gate, controls the signal device to output an entry prohibition signal, or transmits an entry prohibition signal to an external traffic control system. Then, the entry interlock module (540) transmits the control result to the communication execution unit (300).

[0190] The present invention further includes a tunnel image recognition-based emergency communication automatic interlocking method through the above-described embodiment. This is explained with reference to FIG. 8.

[0191] FIG. 8 is a flowchart illustrating a tunnel image recognition-based emergency communication automatic interlocking method according to the present invention.

[0192] Referring to FIG. 8, the present invention includes a step S100 of collecting images from a camera (11), a step S200 of tracking an object, a step S300 of detecting an event by analyzing the images, a step S400 of calculating a risk index and a risk level, a step S500 of calculating a range corresponding interval, a step S600 of generating a control data packet by selecting a corresponding scenario, a step S700 of determining the execution order of the control data packet, a step S800 of adjusting the priority, bandwidth, or transmission path based on network control request data packets and communication network resource status information, and a step S900 of correcting the broadcast output condition, risk index, or entry restriction control by reflecting the diffusion information, acoustic condition, or ventilation condition.

[0193] Step S100 is a step in which the emergency response control terminal (10) collects video from the camera (11). Here, the video collection module (110) of the emergency response control terminal (10) receives a real-time video stream from a plurality of CCTVs in the tunnel and adds camera (11) identifier, shooting time, and tunnel distance coordinate information to each frame.

[0194] The image acquisition module (110) divides the received raw image into frames and performs image distortion correction and brightness normalization. For example, the image acquisition module (110) receives 30 frames of image per second and stores an image with a resolution of 1920×1080 in an analysis buffer. Then, the image acquisition module (110) transmits the generated input image stream to the object tracking module (120).

[0195] Step S200 is a step in which the emergency response control terminal (10) identifies an object in the video and calculates movement speed and trajectory information. Here, the object tracking module (120) of the emergency response control terminal (10) detects a vehicle or person object in the input video stream and assigns a unique object identifier to each object.

[0196] Then, the object tracking module (120) calculates a movement vector and a speed value based on the position change between consecutive frames. For example, the object tracking module (120) marks a specific vehicle's movement speed as a candidate for a sudden stop event when it drops from 80 km / h to 0 km / h. Subsequently, the object tracking module (120) transmits the calculated object location coordinates and movement speed information to the event recognition module (130).

[0197] Step S300 is a step in which the emergency response control terminal (10) determines whether an event has occurred and a reliability value based on object movement information. The event recognition module (130) combines object movement information and image feature information to determine whether the vehicle has stopped, is driving in reverse, or whether smoke or flames have occurred.

[0198] The event recognition module (130) calculates the event reliability as a probability value in the range of 0 to 1. For example, if the vehicle stop duration is 10 seconds or more and the reliability is 0.85 or higher, the event recognition module (130) confirms that an event has occurred. The event recognition module (130) transmits the event type, occurrence location, and reliability value to the risk calculation module (140).

[0199] Step S400 is a step in which the emergency response control terminal (10) calculates a risk index and a risk level based on event information. Here, the risk calculation module (140) of the emergency response control terminal (10) calculates the risk index by reflecting the event type, reliability, duration, number of objects, and traffic density.

[0200] At this time, the risk calculation module (140) normalizes the risk index to a range of 0 to 100. For example, when conditions such as a fire event (weight 40), reliability 0.9, and traffic density 1.2 times are applied, the risk index is calculated as 78. Then, the risk calculation module (140) classifies and calculates the risk level into three stages (caution, warning, emergency) and transmits them to the section management unit (200) and the communication execution unit (300).

[0201] Step S500 is a step in which the emergency response control terminal (10) calculates the response target section using the location of the risk occurrence and the spread information. Here, the range determination module (240) of the emergency response control terminal (10) receives the location of the risk occurrence and the spread prediction information, and determines the response target section based on the tunnel section division information. For example, if the expected spread distance is 300m and the occurrence point is 600m, the range determination module (240) sets the 600m to 900m section as the response section. Then, the range determination module (240) transmits the section identifier list and coordinate information to the communication execution unit (300).

[0202] Step S600 is a step in which the emergency response control terminal (10) selects a response policy according to the event type and the risk level. The scenario module (310) of the emergency response control terminal (10) queries a predefined response policy corresponding to the combination of the event type and the risk level. For example, in the case of the fire + emergency level, the scenario “emergency broadcast transmission + tunnel entry blocking + fire notification” is selected. The scenario module (310) generates a control data packet and transmits it to the stage control module (350).

[0203] Step S700 is a step in which the emergency response control terminal (10) performs communication control according to the selected response policy. The step control module (350) of the emergency response control terminal (10) controls the execution of control data packets sequentially. The step control module (350) transmits a broadcast transmission signal to the broadcast control module (320) and transmits an emergency call connection signal to the telephone control module (330). Each module returns the execution result to the step control module (350).

[0204] Step S800 is a step in which the emergency response control terminal (10) compares the required bandwidth value with network resource status information to reset the bandwidth to be allocated to emergency control data, adjusts the priority of the transmission queue, and switches the transmission session to a backup path if the delay time or packet loss rate exceeds a standard. Here, the status monitoring module (440) collects information on current bandwidth usage, port occupancy, average delay time, and packet loss rate from a switch or router and generates network resource status information.

[0205] The stage control module (350) generates a network control request data packet including a required bandwidth value and a priority level according to the response policy selected by the scenario module (310) and transmits it to the network control unit (400).

[0206] First, the control module (410) receives priority level information and relocates emergency control data packets in the transmission queue to a higher queue than general data packets. For example, the control module (410) sets the emergency broadcast control packet to queue level 7 and the general video data packet to queue level 3.

[0207] The band control module (420) compares the required bandwidth value received from the stage control module (350) with the network resource status information generated by the status monitoring module (440) and calculates the bandwidth to be allocated to emergency control data. For example, if the total link capacity is 1 Gbps and the current usage is 800 Mbps, the band control module (420) limits the normal data transmission rate to 500 Mbps and reallocates 300 Mbps to emergency control data.

[0208] The route switching module (430) switches the transmission session to a backup optical path or backup IP path for the same destination if the average delay time collected by the state monitoring module (440) exceeds 100ms or the packet loss rate exceeds 3%. For example, the route switching module (430) switches the transmission session to the λ2 path if the delay time of the main path λ1 is measured to be 150ms.

[0209] In this way, the network control unit (400) performs priority adjustment, bandwidth reallocation, and transmission path switching to control emergency control data so that it satisfies the set delay time standard (e.g., 50ms or less) and loss rate standard (e.g., 1% or less).

[0210] Step S900 is a step in which the emergency response control terminal (10) corrects the risk index or broadcasting conditions by reflecting environmental conditions. Here, the environmental correction unit (500) of the emergency response control terminal (10) receives the risk index and event occurrence location information received from the risk calculation module (140), and corrects the risk response conditions by reflecting the tunnel internal wind speed information, airflow direction information, ventilation equipment operation status information, and background noise information received from the environmental sensor.

[0211] More specifically, the diffusion prediction module (510) of the environment correction unit (500) receives event occurrence location information received from the risk calculation module (140), wind speed information and airflow direction information received from the environment sensor, and calculates the risk diffusion direction and diffusion range coordinates by reflecting the set prediction conditions. The diffusion prediction module (510) of the environment correction unit (500) transmits the calculated diffusion range coordinate information to the range determination module (240).

[0212] The acoustic correction module (520) of the environment correction unit (500) receives background noise information from an environment sensor, calculates a target sound pressure based on a set range in which audibility is secured relative to the background noise, and transmits broadcast output correction information to the broadcast control module (320).

[0213] The risk correction module (530) of the environment correction unit (500) receives the basic risk index, ventilation equipment operation status information, and traffic density information received from the risk calculation module (140), calculates an environment correction coefficient, and applies it to the basic risk index to calculate a corrected risk index, and transmits the calculated corrected risk index to the risk calculation module (140) and the stage control module (350).

[0214] The risk calculation module (140) compares the correction risk index received from the risk correction module (530) of the environment correction unit (500) with the set risk level criteria to re-determine the risk level, and if a level change is required, transmits a level adjustment signal to the level control module (350).

[0215] The entry linkage module (540) of the environment correction unit (500) receives the diffusion range information calculated from the diffusion prediction module (510) and the correction risk index calculated from the risk correction module (530), generates an entry restriction control signal by comparing it with the set entry restriction conditions, and transmits the control signal to the tunnel entrance control facility (20) through the communication execution unit (300).

[0216] The tunnel entrance control equipment (20) performs at least one of lowering the barrier, closing the automatic barrier gate, displaying a variable information sign warning, illuminating the warning light, transmitting voice from the entrance speaker, or switching the external traffic signal to red based on an entry restriction control signal received from the entry interlocking module (540).

[0218] The present invention is not limited to one specific preferred embodiment, and it is obvious that anyone with ordinary knowledge in the art to which the invention pertains can make various modifications without departing from the essence of the invention as claimed in the claims, and that such modifications will be within the scope of the claims. Explanation of the symbols

[0220] 10: Emergency response control terminal 11: Camera 12: Emergency Terminal 20: Tunnel Entrance Control Equipment 30: External public agency system 40: External transportation system 100: Risk Analysis Unit 110: Image Acquisition Module 120: Object Tracking Module 130: Event Recognition Module 140: Risk Calculation Module 200: Section Management Unit 210: Structure storage module 220: Section partitioning module 230: Direction Analysis Module 240: Range Determination Module 300: Communication Execution Unit 310: Scenario Module 320: Broadcast control module 330: Telephone control module 340: Notification control module 350: Step control module 400: Network control unit 410: Priority control module 420: Band control module 430: Route switching module 440: Status Monitoring Module 450: Emergency Channel Creation Module 460: Predictive Resource Reservation Module 470: Independent Verification Module 480: Section weight adjustment module 500: Environment correction unit 510: Diffusion Prediction Module 520: Acoustic Correction Module 530: Risk Correction Module 540: Entry Interlock Module

Claims

Claim 1 In a tunnel image recognition-based emergency communication automatic interlocking system that recognizes events such as vehicle stopping, driving in reverse, fire, or smoke spread occurring within a tunnel based on video and automatically interlocks and controls emergency communication equipment, the system comprises: a plurality of cameras installed inside or outside the tunnel to capture images of the tunnel situation and transmit the corresponding images to an emergency response control terminal (10); a risk analysis unit (100) that analyzes video data received from the cameras (11) to calculate event type information, event occurrence coordinate information, reliability value, risk index, and risk level information, and calculates the calculated risk index and risk level information; a section management unit (200) that receives the risk index, risk level information, and event occurrence coordinate information transmitted from the risk analysis unit (100), divides the tunnel into a plurality of sections to map the event occurrence location to a section identifier, sets a response target section based on the risk level and expected spread range, and calculates a list of set response target section identifiers; and receives the list of response target section identifiers and risk level information transmitted from the section management unit (200), generates a control data packet based on a response scenario, and automatically activates and calls an emergency terminal (12) mapped to the response target section based on the generated control data packet. A communication execution unit (300) that performs automatic transmission to a configured public institution by establishing a session and transmits notification data including event type information, risk level information, and event occurrence coordinate information to an external public institution system (30) or an external traffic system (40); a network control unit (400) that receives network control request information and communication network resource status information transmitted from the communication execution unit (300), and minimizes transmission delay of emergency control data by adjusting at least one of transmission bandwidth, transmission priority, or transmission path based on the risk level and required transmission conditions;An environment correction unit (500) that calculates the estimated distance and time of diffusion by combining the risk analysis results and environmental sensor information, transmits the calculated estimated distance and time of diffusion to the section management unit (200) to be reflected in the setting of the corresponding target section, calculates a correction risk index and reflects it in the control judgment of the communication execution unit (300), and transmits an entry restriction control signal to at least one of the tunnel entrance control facility (20) and the external traffic system (40) when the correction risk index exceeds the set standard; The emergency terminal (12) is a communication terminal installed inside a tunnel and is configured to automatically open a call channel based on an activation signal received from the communication execution unit (300), transmit status information including terminal identifier information and installation location coordinate information to an external traffic system or a set public institution, and transmit and receive voice data during the call session maintenance time; the environment correction unit (500) calculates a risk spread path, estimated spread distance, and estimated spread time by combining event occurrence location information and event intensity information transmitted from the risk analysis unit (100) with tunnel internal wind speed and airflow direction information received from an environment sensor, transmits the calculated estimated spread distance and estimated spread time to the section management unit (200) to reflect them in the setting of the corresponding target section, and transmits them to the prediction resource reservation module (460) to reflect them in the calculation of the estimated transmission resource; the spread prediction module (510) adjusts the emergency broadcast sound pressure level and output amplification rate based on tunnel cross-section structure information and reflection coefficient information stored in the structure storage module (210) and background noise level information received from the environment sensor, and transmits the adjusted sound output setting value to the communication execution unit (300). A sound correction module (520) that transmits to reflect the broadcast transmission conditions;A tunnel image recognition-based emergency communication automatic interlocking system characterized by comprising: a risk correction module (530) that calculates a risk index correction coefficient by combining the spread-predicted distance and spread-predicted time transmitted from the spread-predicted module (510) with tunnel internal traffic density information and ventilation status information, recalculates a corrected risk index by applying the correction coefficient to an existing risk index with weights, and transmits the recalculated corrected risk index to a communication execution unit (300) to be reflected in the control step judgment; and an entry interlocking module (540) that determines whether to restrict tunnel entry by combining the spread-predicted distance transmitted from the spread-predicted module (510) and the corrected risk index transmitted from the risk correction module (530), and transmits an entry restriction control signal to a tunnel entrance control facility (20) or an external traffic system (40) and transmits control result information to a communication execution unit (300). Claim 2 A tunnel image recognition-based emergency communication automatic interlocking system according to claim 1, wherein the risk analysis unit (100) calculates a risk index by normalizing and weighting at least one of a vehicle speed reduction rate, a stopping duration, a smoke area increase rate, or a flame brightness change rate, and classifies the calculated risk index into multiple stages to classify it into a warning stage or an emergency stage. Claim 3 The tunnel image recognition-based emergency communication automatic interlocking system according to claim 1, wherein the section management unit (200) maps event occurrence location information to a tunnel distance coordinate system, analyzes a vehicle travel direction vector to distinguish between a front section and a rear section, and is configured to additionally include adjacent sections or multiple front sections in the corresponding target sections according to the danger level. Claim 4 A tunnel image recognition-based emergency communication automatic interlocking system, characterized in that, in claim 1, the section management unit (200) dynamically expands the corresponding target section based on the expected spreading distance and expected spreading time, calculates a plurality of section identifiers included in the expected spreading range, and transmits them to the communication execution unit (300). Claim 5 A tunnel image recognition-based emergency communication automatic interlocking system according to claim 1, characterized in that the communication execution unit (300) executes emergency broadcast transmission, automatic emergency terminal activation, and notification to an external public institution system (30) in stages when the danger stage is classified as an emergency stage, and is configured to prioritize the execution of only the emergency broadcast when the danger stage is classified as a warning stage. Claim 6 A tunnel image recognition-based emergency communication automatic interlocking system according to claim 1, characterized in that the emergency terminal (12) is configured to automatically make a call to a control center connection number based on an activation signal received from a communication execution unit (300), and to terminate the session when the call session maintenance time elapses or a call termination signal is received. Claim 7 In claim 1, the network control unit (400) comprises: a priority control module (410) that resets the transmission priority by placing emergency control data packets in the upper queue among a plurality of transmission queues and moving general data packets to the lower queue based on data type code, required bandwidth value, priority level, and maximum allowable delay time information included in network control request information received from the communication execution unit (300); a bandwidth control module that allocates priority bandwidth to emergency control data by comparing the required bandwidth value with communication network resource status information received from the status monitoring module (440) to limit the transmission rate of general traffic, redistribute transmission queue resources, or reset virtual network resources; a path switching module (430) that compares the average delay time and packet loss rate of the main transmission path with the maximum allowable delay time or reference loss rate, and switches the transmission session to a reserve transmission path if the average delay time or packet loss rate exceeds the reference; and a status monitoring module that generates communication network resource status information by collecting transmission bandwidth usage, average delay time, packet loss rate, and port occupancy information from network equipment and transmits it to the priority control module (410), bandwidth control module (420), and path switching module (430). Module (440); an emergency channel creation module (450) that creates a dedicated virtual transmission channel separated from existing communication traffic and sets up emergency control data packets to be transmitted through the dedicated channel when the danger level is determined to be an emergency level; and a predicted resource reservation module (460) that calculates an estimated data transmission amount corresponding to additional broadcast transmission or external notification expected during the spreading period based on the estimated spreading distance and estimated spreading time, and preemptively reserves a bandwidth corresponding to the calculated transmission amount to separate it from general traffic.A tunnel image recognition-based emergency communication automatic interlocking system characterized by comprising: an independent verification module (470) that compares primary communication network status information received from a status monitoring module (440) with secondary communication network status information calculated through multiple actual measurements on the same transmission path, calculates an average value, calculates a standard deviation, and removes outliers to determine whether there is a communication network abnormality if the difference between the two pieces of information exceeds a set standard, and when a communication network abnormality is confirmed, transmits a switching signal to a path switching module (430) and transmits a network reliability degradation index to a risk analysis unit (100); and a section weighting adjustment module (480) that receives section-specific accident history information and traffic density information to calculate section-specific weighting coefficients, reflects the calculated weighting coefficients in the priority level and bandwidth allocation ratio of emergency control data packets, and transmits a section risk accumulation index to a risk analysis unit (100) to reflect it in the recalculation of the risk index when an increase in repetitive accident history is detected. Claim 8 delete

Images