Intelligent monitoring system for integrated detection of complex disasters and road icing

Abstract

The present invention relates to an intelligent monitoring system for integrated detection of complex disasters and road icing, and more specifically, to an intelligent monitoring system for integrated detection of complex disasters and road icing that hierarchically cross-verifies the occurrence of road black ice and disaster situations through an edge artificial intelligence processor based on multi-sensor data collected from road surface and atmospheric environment sensors, fire detection sensors, motion sensors, etc., and immediately transmits a lightweight text warning code via a low-power wide-area network (LPWAN) instead of heavy original video data when a dangerous situation is identified, thereby enabling operation while minimizing power consumption of a single battery even in extreme winter environments. Through this, fire detection, structural displacement monitoring, and road surface icing prediction functions are integrated into a single device, thereby preventing redundant investment. Furthermore, it enables the realization of an advanced integrated control network that operates in an ultra-low power standby mode during normal times and immediately generates an interrupt to perform an organic, top-priority response in the event of a complex disaster.

Description

Technology Field

[0001] The present invention relates to an intelligent monitoring system for integrated detection of complex disasters and road icing, and more specifically, to an intelligent monitoring system for integrated detection of complex disasters and road icing that hierarchically cross-verifies the occurrence of road black ice and disaster situations through an edge artificial intelligence processor based on multi-sensor data collected from road surface and atmospheric environment sensors, fire detection sensors, motion sensors, etc., and immediately transmits a lightweight text warning code via a low-power wide-area network (LPWAN) instead of heavy original video data when a dangerous situation is identified, thereby enabling operation while minimizing power consumption of a single battery even in extreme winter environments. Background Technology

[0003] With the recent acceleration of climate change leading to more frequent localized heavy snowfall and sharp temperature drops during the winter, major chain collisions caused by black ice on roads are emerging as a serious social issue.

[0004] Black ice is a phenomenon in which snow or rain seeping into fine cracks in the asphalt surface freezes into a thin, transparent layer, making it virtually impossible to distinguish from a wet road surface with the naked eye or a standard visible light camera.

[0005] To address this, competent authorities, such as the Korea Expressway Corporation, are monitoring road surface conditions by installing temperature and humidity sensors in major freezing-prone sections. However, the method of relying on the threshold of simple environmental sensors has limitations, as it fails to accurately reflect complex road thermodynamic changes such as atmospheric moisture condensation, melting and refreezing after the application of de-icing agents, resulting in a high false positive rate.

[0006] Furthermore, the Korea Forest Service and local governments have established and are operating separate monitoring systems to monitor various disaster situations, such as forest fires, landslides, and fires inside tunnels.

[0007] The establishment of such individual systems not only results in massive budget duplication and maintenance costs but also hinders the integrated control of disaster information. In particular, disaster-vulnerable areas such as mountain roads and tunnel entrances face a mixture of risks including road icing, wildfires, and rockfalls; however, the lack of a convergence solution capable of comprehensively monitoring and responding to these conditions within a single system has long been pointed out.

[0008] Attempts to introduce high-resolution thermal imaging cameras and artificial intelligence image analysis technology to solve the aforementioned problems are ongoing, but these also face critical hardware challenges in the harsh operating environment of outdoor roads.

[0009] Most road monitoring terminals are located in remote areas where wired power grid installation is impossible, so they must rely on solar panels or batteries and utilize low-power wide-area networks (LPWAN, e.g., LTE Cat.M1) instead of broadband networks.

[0010] Conventional centralized (Cloud AI) systems must continuously transmit high-capacity raw thermal image data to a server, which causes massive communication charges and power consumption.

[0011] A more critical problem is that in sub-zero temperatures, the internal resistance of batteries, such as lithium-ion batteries, increases rapidly, causing efficiency to drop drastically.

[0012] Under these conditions, the instantaneous peak current generated when the communication module is activated to transmit video data causes a rapid voltage drop in the battery, ultimately resulting in the shutdown of the entire system. To prevent this, conventional technologies have required complex hardware configurations, such as cross-charging and discharging multiple batteries or branching power; however, this has not been an effective alternative due to the increased size of the terminal and higher mass production costs. Therefore, there is an urgent need for a groundbreaking paradigm shift in data processing and power control that can ensure stable long-term operation with a single battery in extreme environments while simultaneously monitoring multiple disasters and freezing conditions. Prior art literature

[0014] Korean Registered Patent No. 10-2244242 The problem to be solved

[0015] The main objective of the present invention is to overcome the inefficiency of the conventional system, in which monitoring of various disasters such as wildfires and facility collapses, and winter road icing prediction were operated as separate systems, and to establish a convergence monitoring system that performs these tasks integrally on a single terminal.

[0016] It is intended to enable rapid and comprehensive three-dimensional situational awareness by unifying observation equipment scattered across disaster-prone areas into a single system, thereby eliminating space constraints and unifying the data collection path on the control server side.

[0017] In addition, the present invention aims to fundamentally solve the problem of massive power consumption that occurs when transmitting high-capacity video data using a low-power wide-area network (LPWAN) in remote outdoor areas.

[0018] To achieve this, we aim to drastically reduce energy consumption in communication by shifting the data processing paradigm from a centralized (Cloud AI) to terminal-internal computation (Edge AI). This is done by inducing the terminal to complete internal determination and transmit only a lightweight text warning code when a sensor detects an anomaly, instead of sending the captured video externally.

[0019] Another objective of the present invention is to structurally protect against fatal voltage drops and system shutdowns caused by peak currents of communication modems when operating a system with a single battery in harsh sub-zero winter environments.

[0020] Instead of omitting the conventional complex multiple battery switching or power distribution circuits, an active preheating control logic is introduced that measures the surface temperature of the battery immediately before data transmission and operates the thermal module using its own power, thereby ensuring uninterrupted and stable communication survivability even in extreme cold.

[0021] Finally, the present invention aims to overcome the limitations of existing methods, which frequently resulted in false positives due to determining freezing based on whether a simple environmental sensor threshold was reached, and to maximize the precision of black ice detection by deeply analyzing thermodynamic changes on the road surface.

[0022] We aim to elevate the reliability of AI-based cross-validation to a new level by tracking the latent heat of solidification stagnation section due to moisture phase change in the time-series waveform of road surface temperature, and by fusion analyzing dew point margins and thermal heterogeneity to clearly subdivide simple wet conditions and lethal black ice. means of solving the problem

[0024] The present invention has the following features to solve the above problem.

[0025] The present invention relates to an intelligent monitoring system for integrated detection of complex disasters and road icing, comprising: a camera for acquiring thermal and visible light images of a monitoring area; a fire detection sensor for detecting smoke or gas; and a motion sensor for detecting impact and tilt changes of a structure to monitor complex disaster situations; a data collection unit for collecting icing factors including a road surface sensor for measuring the temperature and moisture of the road surface and an atmospheric environment sensor for measuring the temperature and humidity of the surrounding atmosphere; a wireless communication unit for transmitting and receiving data to and from an external server via a low-power wide-area network (LPWAN); and a power supply unit including a single battery for system operation and a thermal insulation module for maintaining the temperature of said single battery. and a control unit equipped with an edge artificial intelligence processor that is functionally connected to the data collection unit, wireless communication unit, and power supply unit and performs data processing within the terminal; wherein the control unit normally operates in a standby mode that collects sensor data from the road surface sensor, atmospheric environment sensor, fire detection sensor, and motion sensor at a preset interval, and when a first condition is satisfied in which the measured road surface temperature is below a preset threshold temperature and moisture is detected, the control unit activates the camera to acquire a thermal image of the road surface, and inputs the acquired thermal image of the road surface into a freezing determination model mounted on the edge artificial intelligence processor to analyze the temperature distribution pattern of the road surface, thereby independently performing a second cross-verification of whether black ice has occurred.

[0026] Here, the control unit minimizes power consumption of the single battery by omitting the transmission of the original video data and immediately transmitting lightweight text data containing an event warning code to the external server via the wireless communication unit when it is determined that black ice has occurred as a result of the cross-verification, or when a disaster event is detected through the fire detection sensor or motion sensor.

[0027] In addition, the control unit measures the surface temperature of the single battery only immediately before operating the modem of the wireless communication unit when the event warning code is generated in accordance with the detection of the black ice occurrence or disaster event, and if the measured surface temperature is below a preset communication discharge risk temperature, temporarily puts the data transmission operation of the wireless communication unit on hold and preheats the single battery by operating the insulation module using the power of the single battery, and controls the wireless communication unit to transmit the event warning code after the surface temperature reaches a preset target temperature, thereby preventing system shutdown caused by instantaneous peak current during severe cold.

[0028] In addition, the control unit receives weather forecast data for the corresponding monitoring area from an external weather server via the wireless communication unit, and if a hybrid condition is satisfied in which there is a probability of snowfall or rain in the weather forecast data and the slope of the road surface temperature decrease measured in real-time by the data collection unit is greater than or equal to a preset threshold, the control unit actively changes the sensor data collection cycle in the standby mode to a second cycle that is shortened from the preset cycle to provide forecasting.

[0029] In addition, the above-mentioned freezing detection model is trained using a deep learning algorithm to determine the emissivity and local temperature deviation patterns between a thermal image of a wet road surface in a normal state and a thermal image of a road surface where black ice has occurred, and calculates the probability value of black ice occurrence by analyzing the temperature distribution and morphological continuity with the surrounding area in the thermal image of the road surface acquired from the camera.

[0030] In addition, when the control unit receives a fire signal from the fire detection sensor or a structural displacement signal from the motion sensor while operating in the standby mode, it controls the camera to immediately activate and acquire an image of the corresponding monitoring area, regardless of whether the first condition for detecting road surface icing is satisfied.

[0031] In addition, when the control unit acquires an image through the camera activated according to the fire signal or structural displacement signal, it calculates the fire spread area or the tilt displacement amount of the structure within the image through the edge artificial intelligence processor, and while blocking the transmission of the original image data captured by the camera, transmits a disaster warning code in the form of text mapped with the calculated spread area data or tilt displacement amount data through the wireless communication unit.

[0032] In addition, the control unit includes a slope tracking unit that calculates the 'Cooling Rate,' which is the time-series change amount of the road surface temperature data, in real time. When a 'Latent Heat Stagnation Section' is detected where the Cooling Rate temporarily flattens and converges to 0 (Zero) during the process of the road surface temperature dropping below zero, and immediately thereafter a waveform is recognized where the Cooling Rate rapidly drops by exceeding a preset negative (-) threshold, it is determined to be a rapid freezing state caused by the phase transition of the road surface moisture. In this case, the secondary cross-verification process through the edge artificial intelligence processor is bypassed, and a 'Flash Freezing Warning Code' of the highest risk grade is immediately transmitted through the wireless communication unit to preemptively prevent subsequent accidents.

[0033] In addition, the control unit includes a state classification unit that calculates the dew point based on the ambient temperature and humidity measured by the data collection unit, extracts a thermal heterogeneity index—a local temperature deviation value—from the acquired thermal image, and inputs the difference margin between the road surface temperature and the dew point and the thermal heterogeneity index into a freezing determination model to subdivide and classify the current abnormal road surface condition into one of the following mechanisms: ① 'black ice freezing,' which is transparent thin ice caused by atmospheric moisture condensation; ② 'surface freezing' caused by precipitation; ③ 'slush section,' where ice and moisture are mixed due to the reaction of de-icing agents; and ④ simple 'contamination and wetting.' Depending on the subdivided classification result, if classified as ①, it transmits a top-level warning code; if classified as ②, it transmits a mid-level warning code to request a speed limit warning or the dispatch of a snowplow; and if classified as ③, it requests the re-application of de-icing agents. It transmits a code, and when classified as ④, it immediately returns to standby mode without transmitting data to perform individualized customized control logic that conserves power. Effects of the invention

[0035] The most prominent effect of applying the system of the present invention is that it can drastically reduce the budget for infrastructure construction and maintenance of local governments and road authorities.

[0036] By integrating fire detection, structural displacement monitoring, and road surface icing prediction functions into a single device, redundant investment can be prevented. Furthermore, it is possible to realize an advanced integrated control network that operates in an ultra-low power standby mode during normal times and immediately generates an interrupt to perform an organic, top-priority response in the event of a complex disaster.

[0037] In addition, it provides the effect of 혁신적으로 extending the battery life of outdoor devices through data lightweighting logic utilizing edge AI processors.

[0038] It fundamentally blocks the transmission of heavy thermal and visible light source images that severely occupy communication bandwidth, and drastically reduces power consumption by transmitting only risk codes of tens of bytes that have undergone internal calculations, thereby perfectly guaranteeing long-term unmanned operation without the need for frequent battery replacement personnel or the additional installation of massive solar panels.

[0039] In addition, the conditional battery preheating control technology according to the present invention creates the effect of maximizing the reliability of the report by completely eliminating the risk of shutdown of the communication module even during extreme temperature drops in winter.

[0040] In the critical golden time when data must be transmitted to an external server due to freezing or fire events, the battery attempts communication after recovering its optimal discharge temperature on its own, so the inherent power instability problem of outdoor communication equipment in extreme cold can be overcome most smartly and safely with a single battery structure.

[0041] Finally, the thermodynamic mechanism classification and hierarchical cross-validation algorithms demonstrate a powerful effect in preventing the waste of administrative resources caused by false positives and enabling ultra-precise, customized responses tailored to road conditions.

[0042] It can clearly distinguish between invisible black ice and slush mixed with de-icing agents, automatically issuing optimized response instructions for each situation, such as broadcasting warnings on electronic display boards or deploying additional snowplows. In particular, the preemptive detection of rapid freezing (flash freezing) through phase change gradient analysis provides core social value by preventing major chain collision accidents and saving numerous lives. Brief explanation of the drawing

[0044] FIG. 1 is a block diagram showing the overall configuration of an intelligent integrated monitoring system according to one embodiment of the present invention. FIG. 2 is a flowchart illustrating the road surface freezing detection and hierarchical cross-verification process of a control unit according to one embodiment of the present invention. FIG. 3 is a flowchart illustrating an active preheating control process based on the surface temperature of a single battery according to one embodiment of the present invention. FIG. 4 is a flowchart illustrating the process of actively changing the sensor data collection cycle based on external weather forecast data and the slope of road surface temperature decline according to an embodiment of the present invention. FIG. 5 is a flowchart illustrating a priority processing process based on the detection of a disaster event from a fire detection sensor or a motion sensor according to an embodiment of the present invention. FIG. 6 is an exemplary diagram illustrating a rapid freezing determination process based on the time-series cooling gradient waveform of the road surface temperature and the solidification latent heat stagnation section calculated in the gradient tracking unit according to another embodiment of the present invention. FIG. 7 is a flowchart illustrating the process of classifying abnormal road surface conditions into multiple mechanisms based on dew point margin and thermal heterogeneity index in a condition classification unit according to another embodiment of the present invention. Specific details for implementing the invention

[0045] The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the present invention, and the present invention is defined only by the scope of the claims.

[0046] The terms used herein are for describing the embodiments and are not intended to limit the invention. In this specification, the singular form includes the plural form unless specifically stated otherwise in the text. As used herein, "comprises" and / or "comprising" do not exclude the presence or addition of one or more other components in addition to the components mentioned.

[0047] Throughout the specification, the same reference numerals refer to the same components, and "and / or" includes each of the mentioned components and all combinations of one or more. Although "first," "second," etc. are used to describe various components, these components are not limited by these terms.

[0048] These terms are used merely to distinguish one component from another. Therefore, it goes without saying that the first component mentioned below may be the second component within the technical scope of the present invention.

[0049] Unless otherwise defined, all terms used herein (including technical and scientific terms) may be used in a meaning commonly understood by those skilled in the art to which the present invention pertains. Additionally, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively unless explicitly and specifically defined otherwise.

[0050] Spatially relative terms such as "below," "beneath," "lower," "above," and "upper" may be used to facilitate the description of the relationship between one component and other components as illustrated in the drawings. Spatially relative terms should be understood as encompassing different orientations of components during use or operation, in addition to the orientations depicted in the drawings. For example, if a component depicted in a drawing is inverted, a component described as "below" or "beneath" of another component may be placed "above" of that component. Therefore, the exemplary term "below" may encompass both the lower and upper directions. Components may also be oriented in other directions, and accordingly, spatially relative terms may be interpreted according to the orientation.

[0051] The terms “part” or “module” as used in the specification refer to hardware components such as software, FPGAs, or ASICs, and the “part” or “module” performs certain roles. However, the meaning of “part” or “module” is not limited to software or hardware. The “part” or “module” may be configured to reside in an addressable storage medium or may be configured to run one or more processors.

[0052] Accordingly, as an example, a "part" or "module" includes components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables.

[0053] The functions provided within the components and "parts" or "modules" may be combined into a smaller number of components and "parts" or "modules," or further separated into additional components and "parts" or "modules."

[0054] Unless otherwise defined, all terms used herein (including technical and scientific terms) may be used in a meaning commonly understood by those skilled in the art to which the present invention pertains. Additionally, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively unless explicitly and specifically defined otherwise.

[0055] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings.

[0056] FIG. 1 is a block diagram showing the overall configuration of an intelligent monitoring system (1000) for integrated detection of complex disasters and road icing according to one embodiment of the present invention.

[0057] The core objective of the present invention is to overcome the inefficiency of conventional systems where monitoring of various disasters, such as wildfires and facility collapses, and winter road icing prediction were operated as separate systems, and to establish a convergence monitoring system that performs these tasks integrally on a single terminal.

[0058] Referring to FIG. 1, the intelligent monitoring system (1000) is configured to include a data collection unit (100) that collects multi-sensor data to determine the situation of a monitoring area, and a wireless communication unit (200) that transmits and receives data to and from an external server and an external weather server.

[0059] In addition, the system has a hardware architecture in which a power supply unit (300) dedicated to operating the system, a data collection unit (100), a wireless communication unit (200), and a control unit (400) functionally connected to the power supply unit (300) to direct the entire system are organically combined.

[0060] The data collection unit (100) includes multiple sensor devices integrated into a single system for monitoring complex disaster situations and collecting road surface freezing factors.

[0061] Specifically, the data collection unit (100) is equipped with a camera (110) that acquires thermal and visible light images of the surveillance area.

[0062] This camera (110) collects key visual data to transmit to the control unit (400) thermal image of the road surface acquired to distinguish black ice, a transparent ice layer that is virtually impossible to distinguish from the wet road surface with the naked eye or a standard visible light camera.

[0063] Additionally, the data collection unit (100) monitors complex disaster situations such as wildfires, tunnel fires, rockfalls, and facility collapses by including a fire detection sensor (120) that detects smoke or gas and a motion sensor (130) that detects impact and tilt changes of the structure.

[0064] This physically integrates individual observation equipment that was scattered across disaster-prone areas, such as mountain roads and tunnel entrances, thereby resolving space constraints and unifying the data collection path for the control server.

[0065] In addition, the data collection unit (100) collects freezing factors of the road surface in real time by including a road surface sensor (140) that measures the temperature and moisture of the road surface and an atmospheric environment sensor (150) that measures the temperature and humidity of the surrounding air.

[0066] The basic data collected in this way overcomes the limitations of the existing method that relied on the threshold value of a simple environment sensor, and is used as essential preliminary data for the control unit (400) to determine whether black ice has occurred and to cross-verify hierarchically.

[0067] The wireless communication unit (200) is a module that transmits and receives data to and from an external server via a low-power wide-area communication network (LPWAN).

[0068] Given that most road monitoring terminals are located in remote areas where wired power grid installation is impossible, the present invention adopts a low-power broadband communication network instead of a network requiring high-capacity communication bandwidth to exchange lightweight text warning codes or external weather forecast data, thereby suppressing energy consumed in communication.

[0069] The power supply unit (300) is power control hardware consisting of a single battery (310) for operating the system and a thermal insulation module (320) for maintaining the temperature of the single battery (310).

[0070] The present invention eliminates the complex multiple battery switching or power distribution circuits adopted by conventional technologies to prevent shutdown in sub-zero environments, and integrates the driving power using only a single battery (310), thereby fundamentally solving the problems of increased terminal volume and increased mass production costs.

[0071] The inherent power instability and extremely degraded battery efficiency issues arising from this single battery (310) structure are overcome through the combination of a thermal insulation module (320).

[0072] The thermal insulation module (320) functions as a physical protection device that can maintain the optimal discharge temperature required for the operation of a communication modem by being driven using the power of the single battery (310) itself when the internal resistance of the single battery (310), such as lithium-ion, increases rapidly in a cold environment below freezing.

[0073] The control unit (400) is equipped with an edge artificial intelligence processor that performs data processing within the terminal, and acts as the brain of the system that switches the data processing paradigm from a centralized type to a terminal- 자체 computation type.

[0074] The control unit (400) normally operates in a standby mode that collects sensor data from the road surface sensor (140), atmospheric environment sensor (150), fire detection sensor (120), and motion sensor (130) at a preset interval and manages the power and status of the entire hardware.

[0075] Ultimately, the architecture of the intelligent monitoring system (1000) illustrated in FIG. 1 is a hardware design optimized to ensure long-term stable operation with a single battery (310) while fusing a high-resolution camera (110) and an edge artificial intelligence processor into a single device in harsh outdoor road environments.

[0076] Based on this physical foundation, the control unit (400) independently performs advanced software control logic, such as blocking the transmission of heavy video source files or measuring the battery surface temperature to actively preheat.

[0078] Meanwhile, FIG. 2 is a flowchart showing the road surface freezing detection and hierarchical cross-verification process of the control unit (400) according to one embodiment of the present invention.

[0079] The core of this invention is to overcome the limitations of existing methods, which frequently resulted in false positives due to determining freezing based on whether a simple environmental sensor threshold was reached, and to perform data lightweighting logic to drastically reduce energy consumed in communication.

[0080] The control unit (400) normally operates in standby mode with the camera (110) disabled in order to conserve power for the system.

[0081] In this standby mode, the control unit (400) collects sensor data from the road surface sensor (140), the atmospheric environment sensor (150), the fire detection sensor (120), and the motion sensor (130) at a preset interval and continuously monitors the surrounding environment.

[0082] During the periodic data collection process, the control unit (400) determines whether the first detection, i.e., the first condition, is satisfied based on the data collected from the road surface sensor (140).

[0083] Specifically, the measured road surface temperature drops below a preset critical temperature (e.g., minus 1 degree) and at the same time, it is continuously checked whether moisture is detected on the road surface.

[0084] If the road surface temperature is higher than the critical temperature or no moisture is detected, and thus the first condition above is not satisfied (NO), the system considers that there is no risk of freezing on the current road surface.

[0085] Accordingly, the control unit (400) returns to normal standby mode without performing any additional operations or camera operation, and quietly repeats the collection of sensor data according to a preset period.

[0086] On the other hand, when the above first condition is satisfied (YES), the control unit (400) determines that a primary physical environment in which ice may form on the road surface has been created and immediately activates the camera (110) that was deactivated.

[0087] When the camera (110) is turned on, it immediately acquires a thermal image of the road surface, which is the monitoring area, and transmits it to the control unit (400).

[0088] The acquired thermal image of the road surface is immediately input into an ice detection model equipped in an Edge artificial intelligence processor located inside the control unit (400).

[0089] Conventionally, such original video data was sent to a cloud server via a communication network for analysis, but the present invention adopts an edge computing method that performs data processing within the terminal, thereby fundamentally preventing communication delays and power waste caused by high-capacity video transmission.

[0090] The ice detection model equipped in this edge AI processor is an AI model that has been pre-trained with a deep learning algorithm on the emissivity and local temperature deviation patterns between thermal images of a wet road surface in a normal state and thermal images of a road surface where black ice has actually occurred.

[0091] Therefore, beyond simple temperature verification, the temperature distribution and morphological continuity with the surrounding area within the thermal image obtained from the camera (110) are analyzed in combination to calculate a precise probability value of black ice occurrence.

[0092] The control unit (400) utilizes this freezing detection model to deeply analyze the road surface temperature distribution pattern of the thermal image, thereby performing a second cross-verification process to confirm the first detection result that relied only on the environment sensor.

[0093] Through this, it is possible to independently determine whether black ice that is not visible to the naked eye has actually occurred, and if the determination result confirms that it is not black ice but a simple wet road surface (NO), the camera (110) is deactivated and restored to completely filter out false detections.

[0094] If the cross-verification result determines that black ice has occurred (YES), the control unit (400) boldly omits the transmission of the original data of the heavy video, which occupies a significant amount of communication bandwidth.

[0095] Instead, the terminal internally generates only extremely lightweight text data at the tens of bytes level, including an event warning code to notify the outside of the occurrence of freezing.

[0096] Finally, the event warning code in the form of lightweight text data generated in this way is immediately transmitted to an external server via the wireless communication unit (200).

[0097] In this way, by transmitting only a lightweight warning code instead of the original video, the power consumption of the single battery (310) of the outdoor terminal can be minimized, and by deactivating the camera (110) and returning after transmission is complete, a data processing paradigm is achieved that perfectly ensures long-term unmanned operation even in extreme winter environments.

[0099] Meanwhile, FIG. 3 is a flowchart illustrating an active preheating control process based on the surface temperature of a single battery (310) according to one embodiment of the present invention.

[0100] The outdoor road where the present invention is installed is exposed to a harsh environment with sub-zero temperatures during winter, and in such an environment, when the system is driven by a single battery (310), a fatal voltage drop and system shutdown may occur due to the peak current of the communication modem.

[0101] In sub-zero temperatures and severe cold environments, the internal resistance of batteries, such as lithium-ion batteries, increases rapidly, causing efficiency to drop drastically.

[0102] When the communication module is operated to transmit data in this unstable state, the instantaneous peak current that occurs causes a rapid voltage drop in the battery, which eventually results in the entire system shutting down.

[0103] To structurally defend against this, the control unit (400) initiates active protection logic from the stage (S310) where the occurrence of black ice is confirmed or a disaster event is detected and an event warning code is generated through the preceding processes.

[0104] Even if an event warning code is generated to notify the outside world of a dangerous situation, the system does not blindly transmit it but takes a cautious control approach.

[0105] That is, the control unit (400) performs the step (S320) of measuring the surface temperature of a single battery (310) only immediately before operating the modem of the wireless communication unit (200) to transmit the generated event warning code.

[0106] Instead of constantly consuming power and monitoring the temperature, the power consumption of a single battery (310) is minimized even in extreme environments by measuring the temperature only when the communication modem needs to be turned on.

[0107] When the surface temperature is measured, the control unit (400) undergoes a step (S330) of determining whether the measured surface temperature is below a preset communication discharge risk temperature.

[0108] This serves as a key reference point for mathematically determining whether there is a risk of the aforementioned shutdown phenomenon occurring when the communication modem is operated, given the current state of the battery.

[0109] If the measured surface temperature is above the communication discharge risk temperature (NO), it means that the battery is in a sufficiently safe state to withstand the peak current.

[0110] In this case, the control unit (400) bypasses the unnecessary waiting and warm-up sections and proceeds directly to the step (S370) of activating the wireless communication unit (200) to process the data quickly.

[0111] On the other hand, if the measured surface temperature drops below the preset communication discharge risk temperature and is exposed to extreme cold (YES), the control unit (400) enters the active battery preheating protection section.

[0112] First, the control unit (400) performs a step (S340) of temporarily waiting for the data transmission operation of the wireless communication unit (200) to fundamentally block the occurrence of the peak current itself.

[0113] In a safe state where communication is in standby, the control unit (400) performs a self-preheating step (S350) that operates the insulation module (320) using the power of a single battery (310).

[0114] Instead of omitting the conventional complex multiple battery switching or power distribution circuit, an active preheating control logic is introduced to operate the insulation module (320) using the battery's own power.

[0115] While the thermal insulation module (320) is in operation, the control unit (400) undergoes a step (S360) of continuously determining whether the measured surface temperature has reached a preset target temperature.

[0116] If the target temperature has not yet been reached (NO), control is performed to continuously preheat the battery by maintaining the operation phase (S350) of the insulation module (320) again.

[0117] Finally, after the surface temperature of the single battery (310) reaches the target temperature (YES), the control unit (400) performs the step (S370) of finally activating the wireless communication unit (200) that had been temporarily waiting.

[0118] Next, by controlling the completion of the step (S380) of transmitting an event warning code to an external server, system shutdown caused by instantaneous peak current during severe cold is completely prevented.

[0120] Meanwhile, FIG. 4 is a flowchart illustrating the process of actively changing the sensor data collection cycle based on external weather forecast data and the slope of road surface temperature decline according to an embodiment of the present invention.

[0121] Conventional monitoring systems simply collect data by operating sensors according to a pre-set fixed cycle, so they had limitations in that they could not detect the timing of freezing early in extreme conditions with rapidly changing weather, or conversely, waste unnecessary battery power during normal operation.

[0122] To overcome this, the control unit (400) starts control in a normal standby mode (S410) as a basic state for efficiently managing the power of the system.

[0123] In this normal standby mode, the control unit (400) collects sensor data at a preset first cycle and monitors the surrounding environment of the monitoring area.

[0124] Here, the first cycle refers to a basic observation interval set generously to minimize single battery consumption in normal conditions where the environment is clear or the risk of freezing is low.

[0125] The system of the present invention does not passively rely solely on local current sensor measurements around the terminal, but actively utilizes external macroscopic weather information to preemptively recognize impending risks.

[0126] To this end, the control unit (400) performs the step (S420) of receiving weather forecast data for the corresponding monitoring area from an external weather server via the wireless communication unit (200).

[0127] At the same time, the system closely tracks microscopic physical changes in the local area where the terminal is installed without missing any.

[0128] The control unit (400) performs the step (S430) of calculating the slope of the road surface temperature drop based on the road surface temperature data measured in real time by the data collection unit (100).

[0129] This is an advanced process that calculates not just the simple absolute value of the current temperature, but the time-series rate of temperature change indicating how steeply the road surface temperature is dropping over time.

[0130] The control unit (400), having secured both external macroscopic weather forecasts and internal microscopic measurement data, enters a step (S440) of determining whether a hybrid condition is satisfied by fusing these two different pieces of information to determine the imminent risk of freezing.

[0131] This hybrid condition serves as a key judgment criterion that enables the present invention to significantly reduce false weather detections and enable preemptive surveillance and response before the formation of black ice.

[0132] Specifically, this hybrid condition is established only when two detailed conditions are simultaneously satisfied.

[0133] The first is an external condition in which there is a physical possibility of moisture being supplied to the road due to the existence of a probability of snowfall or rain in the received weather forecast data, and the second is an internal condition in which the slope of the real-time calculated road surface temperature is dropping steeply above a preset threshold.

[0134] If the weather forecast predicts snow or rain and at the same time the actual road temperature drops rapidly, satisfying the above hybrid condition (YES), the control unit (400) determines that there is an imminent danger condition in which black ice or fatal road icing may occur.

[0135] Accordingly, the control unit (400) immediately performs the step (S450) of actively changing the sensor data collection cycle in standby mode to a second cycle that is shortened from the existing first cycle.

[0136] In this way, if the sensor data collection cycle is actively changed to a dense second cycle, the system has the effect of switching to an emergency mode that significantly strengthens ice prediction.

[0137] Since the data measurement interval is drastically shortened, road conditions that rapidly freeze due to snow or rain can be captured in near real-time, and a full defensive posture is established to send an immediate warning code to the server without missing the golden time for freezing.

[0138] On the other hand, if there is no snow or rain forecast at all, or if the road surface temperature remains stable and neither of the above hybrid conditions is satisfied (NO), the system determines that the current situation is relatively calm with a low probability of rapid road icing.

[0139] At this time, since there is no need to waste limited power by operating the sensor more frequently, the system enters the step (S460) of maintaining the existing sensor data collection cycle, the first cycle, and continues the standby mode.

[0140] Consequently, through such weather data-linked hybrid forecasting cycle change logic, the present invention maximizes detection precision by increasing the sensor operating rate when risk is high, and thoroughly conserves power by maintaining a sufficient collection cycle during normal times.

[0141] This demonstrates exceptional technical effectiveness by flexibly overcoming the critical power limitations of monitoring systems that must survive for extended periods in harsh outdoor environments using only a single battery, through smart software control.

[0143] Meanwhile, FIG. 5 is a flowchart showing a priority processing process based on the detection of a disaster event from a fire detection sensor (120) or a motion sensor (130) according to one embodiment of the present invention.

[0144] The present invention is characterized by integrated monitoring of critical complex disaster situations, such as fires or facility collapses as well as winter road icing, from a single terminal, and establishing an interrupt system that responds with top priority by disregarding the general control sequence when an emergency occurs.

[0145] The control unit (400) manages the basic state of the entire system while maintaining a normal standby mode (S510) to minimize power consumption.

[0146] In this standby mode, the control unit (400) collects data from each sensor according to a preset period and quietly monitors for abnormalities in the monitoring area.

[0147] During this normal monitoring, the control unit (400) performs a step (S520) of determining whether a fire signal from the fire detection sensor (120) or a structural displacement signal from the motion sensor (130) is received.

[0148] This is a key surveillance gateway for the system to immediately detect whether urgent and lethal disaster events, such as wildfires, tunnel fires, or rockfalls and collapses, have occurred within the surveillance area.

[0149] If no fire signal or structural displacement signal is received (NO), the monitoring area is in a calm state from a disaster perspective, so the control unit (400) returns to the normal standby mode operation step (S510) to prevent unnecessary power consumption or continues to perform the existing freezing detection logic.

[0150] As a result, when not in an emergency, the system thoroughly maintains its primary mission of standby and low-power states.

[0151] On the other hand, if a fire or displacement signal is received and a disaster event is detected (YES), the control unit (400) performs the step (S530) of immediately generating a priority interrupt regardless of whether the first condition for detecting road surface freezing is satisfied.

[0152] In other words, it completely suspends general ice determination logic, such as checking whether the current road surface temperature is below freezing or if there is moisture, and elevates disaster situations directly linked to casualties to the system's top priority.

[0153] As a priority interrupt occurs, the control unit (400) immediately activates the disabled camera (110) and rapidly proceeds to the step (S540) of acquiring an image of the surveillance area.

[0154] As soon as a disaster signal is received, the camera (110) is immediately turned on to secure visual data of the scene first without time-series delay, thereby establishing a data basis for accurate situation assessment within the golden time.

[0155] When an image is acquired, the control unit (400) does not immediately send data to the server, but enters a data blocking and edge calculation step (S550) in which it calculates the fire spread area or the tilt displacement amount of the structure within the image through an edge artificial intelligence processor.

[0156] This is an advanced, independent computational process in which artificial intelligence directly quantifies and calculates the scale and severity of a disaster at the site (Edge) where the terminal is installed.

[0157] At the same time, the control unit (400) thoroughly blocks the transmission of original video data, which causes an overload of the communication network and massive power waste.

[0158] Due to the characteristics of an outdoor terminal using a low-power wideband network (LPWAN), transmitting a high-capacity original video can cause the battery to be rapidly depleted; therefore, after the edge computation is completed, the original data itself is prevented from going outside to maximize the survivability of the single battery (310).

[0159] When the edge calculation and original data blocking are completed, the control unit (400) performs the step (S560) of generating a disaster warning code in the form of text in which the self-calculated fire spread area or the tilt displacement amount data of the structure is numerically mapped.

[0160] The weight of data is 혁신적으로 reduced by replacing heavy video files with lightweight text code symbols of tens of bytes.

[0161] Finally, the control unit (400) completes the step (S570) of immediately transmitting the disaster warning code in the form of text to an external server via the wireless communication unit (200).

[0162] As a result, through this series of interrupt and edge operation logic, the present invention realizes an advanced convergence response system capable of reporting the precise scale of a disaster to a control server as a top priority without any delay or battery shutdown when various types of disaster situations occur.

[0164] Meanwhile, FIG. 6 is an example diagram showing a rapid freezing determination process based on the time-series cooling gradient waveform of the road surface temperature and the solidification latent heat stagnation section calculated by the gradient tracking unit (410) according to another embodiment of the present invention.

[0165] This invention overcomes the limitations of conventional monitoring systems that determine freezing based solely on sub-zero temperatures and presents an advanced algorithm in which artificial intelligence directly tracks the thermodynamic physical laws that occur when moisture undergoes a phase transition into ice.

[0166] When moisture on a road freezes, the temperature does not drop in a straight line; instead, it is necessarily accompanied by a physical phenomenon in which heat is released to the outside as the state changes from liquid to solid.

[0167] Accordingly, the slope tracking unit (410) provided within the control unit (400) does not merely check the absolute value of the current temperature but performs a pivotal role in calculating the 'Cooling Rate,' which is the amount of change in the time series of road surface temperature data, in real time.

[0168] Looking at the waveform graph in Fig. 6, the horizontal axis X-axis represents the flow of time (t), and the vertical axis Y-axis represents the measured road surface temperature (T).

[0169] The slope tracking unit (410) tracks the mathematical ratio of the change in temperature relative to the change in time in real time in this two-dimensional time series space and analyzes the shape of the temperature waveform.

[0170] First, 'Section A', the initial stage where the road surface temperature begins to drop rapidly below freezing, is defined as a simple cooling section.

[0171] In section A, the road surface is cooled by the external cold, and the temperature drop slope takes on a negative value less than 0 and drops very steeply, exhibiting a simple cooling pattern.

[0172] Subsequently, as the temperature continues to drop, when it reaches the phase change starting point (t1) located near the freezing point, 0°C critical line, the moisture present on the road surface begins to solidify into ice in earnest.

[0173] At this time, as the moisture changes phase into ice, it releases latent heat of solidification into the atmosphere, and due to this latent heat, the temperature of the road surface temporarily stops dropping and becomes stagnant.

[0174] The slope tracking unit (410) detects this special waveform section, which is the ‘solidification latent heat stagnation section’, as ‘Section B’, which is the ‘solidification latent heat stagnation section’, at this very point, that is, during the process of the road surface temperature dropping below zero.

[0175] The horizontal waveform in section B serves as the most definitive and obvious physical evidence indicating that black ice, a deadly transparent thin ice, is currently forming on the road in real time.

[0176] After this latent heat release section ends and the freezing completion point (t2) where the moisture completely freezes on the road surface is passed, the slope of the waveform changes abruptly again.

[0177] As the phase change is completely finished, the road surface assimilates back to the cold ambient temperature and begins to cool rapidly, and the waveform enters 'Section C', the rapid freezing completion and re-cooling section.

[0178] In this section C, the slope tracking unit (410) finally recognizes a waveform in which the stagnant cooling slope exceeds a preset negative (-) threshold and drops very steeply again.

[0179] When this step-shaped waveform, which follows a steep drop followed by a horizontal stagnation and then a steep drop again, is completed, the control unit (400) determines this as a clear rapid freezing (flash freezing) state caused by a phase change of the road surface moisture.

[0180] When this stage is reached where rapid freezing is confirmed, the speed of freezing progression and the urgency of the risk of accidents reach an extreme, so the control unit (400) makes a bold control decision to bypass even the secondary cross-verification process of the thermal image through the edge artificial intelligence processor.

[0181] It is determined that there is absolutely no need to delay the system's computation time, as the solidification latent heat waveform itself is a much faster and unmanipulable physical confirmation than AI image analysis.

[0182] Finally, the control unit (400) skips the second cross-verification and immediately generates a 'Flash Freezing warning code' corresponding to the highest risk level and immediately transmits it to the control server via the wireless communication unit (200).

[0183] This solidification latent heat tracking logic fundamentally eliminates false positives and demonstrates powerful advancement by preemptively detecting the formation of black ice, which causes major collisions, without any time-series delay, thereby completely preventing subsequent accidents.

[0185] Meanwhile, FIG. 7 is a flowchart illustrating the process of classifying abnormal road surface conditions into multiple mechanisms based on dew point margin and thermal heterogeneity index in a state classification unit (420) according to another embodiment of the present invention.

[0186] This embodiment discloses an advanced artificial intelligence algorithm that overcomes the blind spot of conventional technology, which judged only the presence or absence of freezing using black-and-white logic, and prevents false detections by deeply analyzing complex thermodynamic changes on the road surface, thereby enabling customized administrative responses for each situation.

[0187] The state classification unit (420) provided within the control unit (400) first performs a data collection step (S710) for collecting and calculating essential parameters to diagnose the complex thermodynamic state of the road surface.

[0188] Specifically, based on ambient air temperature and humidity data measured from the data collection unit (100), the dew point at which moisture in the air begins to liquefy and form in the current environment is mathematically calculated with precision.

[0189] Next, the state classification unit (420) derives the ‘Dew Point Margin,’ which is the difference between the measured road surface temperature and the previously calculated dew point.

[0190] At the same time, the thermal image obtained from the thermal imaging camera among the cameras (110) is analyzed to extract a 'Thermal Heterogeneity Index,' which is a local temperature deviation value indicating how uneven the temperature distribution of the road surface within the monitoring area is, thereby completing the preparation for composite determination.

[0191] Based on the collected and calculated indicators, the state classification unit (420) enters a step (S720) of primarily determining whether the dew point margin is below a preset threshold.

[0192] If the dew point margin exceeds a threshold value and it is determined that there is almost no possibility of atmospheric moisture condensation (NO), the flow terminates by going through the step (S730) of classifying the road surface as a dry road surface without freezing risk, and then immediately returning to normal mode without data transmission to conserve power (S740).

[0193] On the other hand, if a thermodynamic condition is established where the dew point margin drops below a critical value and moisture can condense (YES), the state classification unit (420) performs a step (S750) of determining whether the current road surface temperature is below the freezing point of 0°C.

[0194] This stage serves as a critical dividing point that clearly distinguishes whether the road surface is in lethal sub-zero freezing conditions where moisture can freeze, or simply a wet road at above-freezing temperatures.

[0195] In a freezing condition (YES) where the road surface temperature is confirmed to be below the freezing point, the state classification unit (420) undergoes a step (S760) of additionally determining whether the thermal heterogeneity index extracted from the thermal image above is below a preset threshold value.

[0196] The fact that the road surface temperature is below zero and the heat distribution on the surface is very uniform (low heterogeneity index) is the strongest physical evidence suggesting that the asphalt surface was rapidly cooled with a thin and even coating of moisture.

[0197] Accordingly, when the thermal heterogeneity index is below the threshold value (YES), the state classification unit (420) performs the step (S770) of classifying the current road surface into a thin, transparent ice layer caused by atmospheric moisture condensation, i.e., the worst situation with the highest risk of accidents, which is 'black ice (uniform freezing)'.

[0198] This is a unique identification mechanism exclusive to this invention, in which artificial intelligence precisely detects invisible black ice through thermal uniformity rather than simple temperature, as if striking it.

[0199] Conversely, if the road surface temperature is below zero but the thermal heterogeneity index exceeds a threshold value and the road surface temperature is locally very uneven (NO), the state classification unit (420) performs a step (S780) of subdividing and classifying it into a 'partial freezing (frost, rough ice)' state in which rough surface freezing caused by precipitation or de-icing agent reaction residues are mixed.

[0200] Since this is a physical state distinct from thin black ice that is evenly frozen, the level of subsequent response will be different.

[0201] If the condition is classified as a partial freezing state at step S780, the state classification unit (420) generates an intermediate warning code one level lower than the black ice (uniform freezing) state and transmits it to an external server via the wireless communication unit (200).

[0202] This mid-level warning code notifies the control server that a rough layer of ice or frost may partially form on the road surface, which could impair the vehicle's braking performance, and can be linked to administrative responses that trigger speed limit warnings on electronic display boards in the relevant section or requests for snow removal vehicles.

[0203] In this way, the S780 classification result is processed as an independent warning level distinct from the highest warning code of S770, thereby realizing a gradual and precise customized road response system proportional to the level of risk.

[0204] Meanwhile, if the condition of the image where the road surface temperature exceeds the freezing point (NO of S750) is branched, the state classification unit (420) goes through a step (S800) of determining whether the thermal heterogeneity index is greater than or equal to a threshold value.

[0205] If the thermal heterogeneity index is high (YES) despite the temperature of the image, a step (S810) of classifying it into a ‘partially wet (slush, standing water)’ state in which ice and water are mixed messily due to de-icing agent reactions or contaminants is performed, and conversely, if the heterogeneity index is low (NO), a step (S820) of classifying it into a simple ‘uniformly wet (wet road surface)’ state in which the surface is evenly wet due to rainfall, etc. is performed.

[0206] Finally, when the dew point margin and thermal heterogeneity index are fused and input into the freezing determination model to complete the detailed classification into the four mechanisms mentioned above, the state classification unit (420) terminates the process by going through the step (S790) of transmitting a customized warning code mapped to the classification result to an external server via the wireless communication unit (200).

[0207] That is, when classified as S770, a top-level warning code is sent to trigger a warning on the electronic display board; when classified as S780, a middle-level warning code one level lower than the black ice (uniform freezing) condition is generated and transmitted to an external server to trigger a speed limit warning on the electronic display board of the section or a request for snowplow dispatch; and when classified as S810, a maintenance code requesting the re-spraying of de-icing agents by the snowplow is transmitted, thereby completing a customized control logic that is perfectly optimized for each situation.

[0209] Although preferred embodiments of the present invention have been described above, the present invention is not limited to the specific embodiments described above. That is, those skilled in the art to which the present invention pertains can make numerous changes and modifications to the present invention without departing from the spirit and scope of the appended claims, and all such appropriate changes and modifications should be deemed to fall within the scope of the present invention as equivalents. Explanation of the symbols

[0211] 100 : Data Collection Unit 110 : Camera 120: Fire detection sensor 130 : Motion sensor 140: Road surface sensor 150: Air quality sensor 200 : Wireless Communications Unit 300 : Power supply unit 310: Single battery 320 : Thermal insulation module 400 : Control unit 410 : Gradient tracking unit 420 : Status Classification Section 1000: Intelligent Integrated Monitoring System

Claims

Claim 1 An intelligent monitoring system for integrated detection of complex disasters and road icing comprises: a camera (110) for acquiring thermal and visible light images of a monitoring area, a fire detection sensor (120) for detecting smoke or gas, and a motion sensor (130) for detecting impact and tilt changes of a structure to monitor complex disaster situations, and a data collection unit (100) for collecting icing factors including a road surface sensor (140) for measuring the temperature and moisture of the road surface and an atmospheric environment sensor (150) for measuring the temperature and humidity of the surrounding atmosphere; a wireless communication unit (200) for transmitting and receiving data to and from an external server via a low-power wide-area network (LPWAN); and a power supply unit (300) including a single battery (310) for operating the system and a thermal insulation module (320) for maintaining the temperature of the single battery (310). An intelligent monitoring system comprising: a control unit (400) functionally connected to the data collection unit (100), wireless communication unit (200), and power supply unit (300), and equipped with an edge artificial intelligence processor that performs data computation within the terminal; wherein the control unit (400) normally operates in a standby mode that collects sensor data from the road surface sensor (140), atmospheric environment sensor (150), fire detection sensor (120), and motion sensor (130) at a preset interval, and when a first condition is satisfied in which the measured road surface temperature is below a preset threshold temperature and moisture is detected, the camera (110) is activated to acquire a thermal image of the road surface, and the acquired thermal image of the road surface is input into a freezing determination model mounted on the edge artificial intelligence processor to analyze the temperature distribution pattern of the road surface, thereby independently performing a second cross-verification of whether black ice has occurred. Claim 2 An intelligent monitoring system according to claim 1, wherein the control unit (400) minimizes power consumption of the single battery (310) by omitting the transmission of original video data and immediately transmitting lightweight text data including an event warning code to the external server via the wireless communication unit (200) when it is determined that black ice has occurred as a result of the cross-verification, or when a disaster event is detected through the fire detection sensor (120) or motion sensor (130). Claim 3 In paragraph 2, the control unit (400) measures the surface temperature of the single battery (310) only immediately before operating the modem of the wireless communication unit (200) when the event warning code is generated according to the detection of the black ice occurrence or disaster event, and when the measured surface temperature is below a preset communication discharge risk temperature, temporarily puts the data transmission operation of the wireless communication unit (200) on hold and preheats the single battery (310) by operating the insulation module (320) using the power of the single battery (310), and controls the wireless communication unit (200) to transmit the event warning code after the surface temperature reaches a preset target temperature, thereby preventing system shutdown due to instantaneous peak current during severe cold. Claim 4 An intelligent monitoring system according to claim 1, wherein the control unit (400) receives weather forecast data for the corresponding monitoring area from an external weather server via the wireless communication unit (200), and when a hybrid condition is satisfied in which there is a probability of snowfall or rain in the weather forecast data and the slope of the road surface temperature falling in real time measured by the data collection unit (100) is greater than or equal to a preset threshold, the sensor data collection cycle in the standby mode is actively changed to a second cycle that is shortened from the preset cycle to predict. Claim 5 In claim 1, the above-mentioned freezing determination model is an intelligent monitoring system that learns the emissivity and local temperature deviation patterns between a thermal image of a wet road surface in a normal state and a thermal image of a road surface where black ice has occurred using a deep learning algorithm, and calculates a black ice occurrence probability value by analyzing the temperature distribution and morphological continuity with the surrounding area in the thermal image of the road surface acquired from the camera (110). Claim 6 An intelligent monitoring system according to claim 1, wherein the control unit (400) controls the camera (110) to immediately activate and acquire an image of the corresponding monitoring area, regardless of whether the first condition for detecting road surface freezing is satisfied, when a fire signal from the fire detection sensor (120) or a structural displacement signal from the motion sensor (130) is received while operating in the standby mode. Claim 7 In claim 6, the control unit (400), upon acquiring an image through the camera (110) activated according to the fire signal or structural displacement signal, calculates the fire spread area or the amount of tilt displacement of the structure within the image through the edge artificial intelligence processor, and while blocking the transmission of the original image data captured by the camera (110), transmits a disaster warning code in the form of text mapped with the calculated spread area data or tilt displacement data through the wireless communication unit (200), an intelligent monitoring system.

Images