Environment-adaptive-based detector temperature range dynamic switching method and system

Abstract

The application discloses an environment self-adaptive based detector temperature range dynamic switching method and system, relates to the field of general control or regulation system, and comprises the following steps: identifying high-temperature pixel points exceeding a preset temperature threshold in infrared thermal imaging data as a target high-temperature area; calculating a high-temperature area proportion value of the target high-temperature area in the whole picture of the infrared thermal imaging data; determining an area proportion threshold and a duration threshold corresponding to a current scene mode; starting a timer to determine a high-temperature duration when the high-temperature area proportion value exceeds the area proportion threshold; calculating a temperature change rate when the high-temperature duration exceeds the duration threshold; and issuing an instruction to a control circuit of a thermal imaging module when the temperature change rate exceeds a preset change threshold, so that a target detector is switched to a target measurement range to monitor the target high-temperature area. The application can dynamically switch the detector range and ensure the detection accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of general control or regulation systems, and particularly relates to a method and system for dynamically switching temperature ranges of a detector based on environmental adaptation. BACKGROUND

[0002] In emergency rescue sites such as fires and explosions, the environment is often extremely complex and dangerous, full of smoke and high temperatures. In order to protect the lives of rescue personnel (especially firefighters) and efficiently perform search and rescue tasks, visual multifunctional detectors have become indispensable firefighting equipment. Such devices integrate infrared thermal imaging functions and can help firefighters detect heat distribution in smoky or lightless environments, thereby quickly navigating, searching for the vital signs of trapped personnel, or locating hidden fire cores.

[0003] In related technologies, a manual or simple range switching function is provided for the detector. For example, the device is set to work in a low-temperature range (such as -20℃ to 150℃) for searching for human bodies. When the firefighter enters the fire site and determines through the screen that he is approaching an extreme high-temperature area such as the core of the flame, he needs to actively switch the device to a high-temperature measurement range (such as 0℃ to 1200℃) by pressing physical buttons or operating menus. After switching, the detector can measure and present details of the temperature of high-temperature targets such as flames.

[0004] However, the rapidly developing fire will be displayed as a white high-light area on the screen due to exceeding the upper limit of the low-temperature range. If the range is not adjusted in time, the inaccurate identification of the high-temperature area by the detector will affect the firefighter's judgment of the dangerous evolution trend and delay the decision-making opportunity. SUMMARY

[0005] The present application provides a method and system for dynamically switching temperature ranges of a detector based on environmental adaptation, which is used to dynamically switch the range of the detector and ensure detection accuracy.

[0006] In a first aspect, the application provides a method for dynamically switching the temperature range of a detector based on environmental adaptation. The method is applied to a detector control system and includes the following steps: obtaining infrared thermal imaging data, identifying high-temperature pixel points in the infrared thermal imaging data that exceed a preset temperature threshold as a target high-temperature region; calculating the area proportion of the target high-temperature region in the full picture of the infrared thermal imaging data to obtain a high-temperature region proportion value; determining the area proportion threshold and the duration threshold corresponding to the current scene mode in the mode strategy library; when the high-temperature region proportion value exceeds the area proportion threshold, starting a timer and determining the high-temperature duration based on the timer; when the high-temperature duration exceeds the duration threshold, calculating the temperature change rate of the target high-temperature region based on the infrared thermal imaging data of multiple consecutive frames within the high-temperature duration; and when the temperature change rate exceeds a preset change threshold, issuing an instruction to the control circuit of the thermal imaging module to cause the target detector to switch to a target measurement range to monitor the target high-temperature region.

[0007] In the above embodiment, the detector control system identifies the high-temperature region based on the infrared thermal imaging data and dynamically adjusts the detection strategy according to different scene characteristics by calculating the high-temperature region proportion, duration, and temperature change rate to trigger the range switching instruction, ensuring that the detector works in the optimal measurement range in complex fire environment and avoiding the delay and risk caused by manual switching.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the step of obtaining infrared thermal imaging data and identifying high-temperature pixel points in the infrared thermal imaging data that exceed a preset temperature threshold as a target high-temperature region specifically includes the following steps: based on historical posture coordinate data and historical infrared data collected by multiple associated detectors, constructing a three-dimensional scene map containing multiple known heat source positions and known heat source characteristics; the associated detector is another detector that is located in the same monitoring area as the target detector and has the same device function; obtaining infrared thermal imaging data and extracting a high-temperature combined region of all high-temperature pixel points in the infrared thermal imaging data; calculating the estimated position of the combined region in the three-dimensional scene map according to the image features and posture coordinate features of the high-temperature combined region; and based on the estimated position, the known heat source positions, and the known heat source characteristics, determining multiple known heat source regions and a new heat source region as the target high-temperature region.

[0009] In the above embodiment, the detector control system constructs a three-dimensional scene map using historical data of multiple associated detectors, analyzes the image features and posture coordinate features, and accurately locates the new heat source and compares it with the known heat source, which not only improves the accuracy of heat source positioning but also timely discovers abnormal heat sources to facilitate the analysis of fire situation.

[0010] In some embodiments of the first aspect, after the step of determining the plurality of known heat source areas and the new heat source area as the target high-temperature area based on the estimated position, the known heat source positions, and the known heat source features, the method further comprises: obtaining known identity labels of the known heat source areas and binding new identity labels for the new heat source area; binding real-time temperature data, time stamps, and collection device IDs for the known identity labels and the new identity labels, generating data update records, and uploading the data update records to a state record database.

[0011] In the above embodiments, the detector control system assigns unique identity labels to heat source areas and binds real-time temperature, time stamps, and other key information, establishing a complete heat source state tracking mechanism. Through real-time uploading of data update records, information sharing and collaborative monitoring between multiple devices are achieved, improving the response speed and monitoring efficiency of the system to changes in fire field situation.

[0012] In some embodiments of the first aspect, the step of constructing a three-dimensional scene map containing a plurality of known heat source positions and known heat source features based on historical posture coordinate data and historical infrared data collected by a plurality of associated detectors specifically comprises: reading historical posture coordinate data and historical infrared data collected by a plurality of associated detectors from a state record database, constructing a data index table according to collection device IDs and time stamps; performing coordinate system conversion and registration on the historical posture coordinate data according to the time stamp order in the data index table to generate spatial point cloud data in a unified coordinate system with time sequence association; mapping the historical infrared data to the corresponding spatial point cloud positions in the spatial point cloud data, and calculating heat source feature parameters of a plurality of spatial points in the spatial point cloud data; and constructing a three-dimensional scene map containing known heat source positions and known heat source features based on the spatial point cloud data and the heat source feature parameters.

[0013] In the above embodiments, the detector control system realizes unified management and analysis of historical data. Through coordinate system conversion and registration, a time sequence associated point cloud data model is constructed, and heat source feature parameters are mapped, forming a high-precision three-dimensional scene map for fire field data analysis.

[0014] In some embodiments of the first aspect, after the step of reading the historical posture coordinate data and historical infrared data collected by the plurality of associated detectors from the state record database, constructing a data index table according to the collection device ID and the timestamp, the method further comprises: grouping the historical infrared data of each collection device according to the monitoring area based on the data index table, calculating the spatial distribution density of the infrared heat source and the temperature accumulation time length in each monitoring area; identifying the heat source gathering point and the heat diffusion channel in the area according to the spatial distribution density and the temperature accumulation time length, constructing a heat source activity map; superimposing the heat source activity map on the corresponding posture coordinate data in the data index table, establishing a mapping relationship between the heat source activity area and the spatial position; based on the mapping relationship, classifying the spatial area in the three-dimensional scene map according to the heat source activity level, and adjusting the sampling density and the monitoring frequency of the target detector in different areas according to the classification result.

[0015] In the above embodiments, the detector control system can identify the heat source gathering point and the heat diffusion channel, construct a heat source activity map, and classify the area according to the activity level based on the spatial mapping relationship, which realizes the optimization of the sampling strategy of the detector and improves the overall monitoring efficiency of the system.

[0016] In some embodiments of the first aspect, before the step of calculating the temperature change rate of the target high-temperature area according to the continuous multiple frames of infrared thermal imaging data in the high-temperature duration when the high-temperature duration exceeds the duration threshold, the method further comprises: calculating the temperature average value of all pixel points in the current field of view of the detector as a temperature reference value; traversing each pixel point in the current field of view, marking the pixel points with a temperature value higher than a preset multiple of the temperature reference value as high-temperature points; performing connected component analysis on the high-temperature points, merging adjacent high-temperature points into standard high-temperature areas, and assigning a unique identifier to each standard high-temperature area; recording the pixel number, area centroid coordinates and area boundary coordinates of each standard high-temperature area, and constructing a region attribute list.

[0017] In the above embodiments, the detector control system establishes a standardized high-temperature area identification and analysis process, which can accurately identify and classify different high-temperature areas by calculating the temperature reference value, marking the high-temperature points and performing connected component analysis, thereby improving the analysis efficiency of the system.

[0018] In some embodiments of the first aspect, after the step of recording the pixel number, the region centroid coordinates and the region boundary coordinates of each standard high-temperature region, and constructing the region attribute list, the method further comprises: based on the region attribute list, expanding the boundary coordinates of each standard high-temperature region, adding a preset distance of a warning buffer zone outside the original boundary; adding the pixel points of the warning buffer zone to the region attribute list, expanding the region boundary coordinate record, and obtaining an expanded region attribute list; extracting the centroid coordinates of each standard high-temperature region after expansion from the expanded region attribute list, calculating the distance relationship between adjacent regions, and constructing a region distance index; according to the region distance index, identifying a region combination with a distance less than a preset distance threshold, and generating a merging warning information.

[0019] In the above embodiments, the detector control system realizes the early warning expansion function of the high-temperature region, and by adding the warning buffer zone and constructing the region distance index, the diffusion trend of the heat source can be predicted in advance, more comprehensive risk warning information is provided for fire rescue, and the warning accuracy of the system is improved.

[0020] In the second aspect, the embodiments of the present application provide a detector control system, which comprises one or more processors and a memory; the memory is coupled with the one or more processors, and the memory is used to store computer program codes, the computer program codes comprise computer instructions, and the one or more processors invoke the computer instructions to make the detector control system execute the method described in the first aspect and any possible implementation manner of the first aspect.

[0021] In the third aspect, the embodiments of the present application provide a computer program product comprising instructions, when the computer program product is executed on the detector control system, the detector control system executes the method described in the first aspect and any possible implementation manner of the first aspect.

[0022] In the fourth aspect, the embodiments of the present application provide a computer readable storage medium comprising instructions, when the instructions are executed on the detector control system, the detector control system executes the method described in the first aspect and any possible implementation manner of the first aspect.

[0023] It can be understood that the detector control system provided in the second aspect, the computer program product provided in the third aspect and the computer storage medium provided in the fourth aspect are all used to execute the method provided in the embodiments of the present application. Therefore, the beneficial effects that can be achieved are referred to the beneficial effects in the corresponding method, which will not be described here.

[0024] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0025] 1. Due to the adoption of multi-dimensional parameter analysis based on infrared thermal imaging and self-adaptive range switching mechanism, the system can monitor the area proportion, duration and temperature trend of high-temperature area in real time, and automatically adjust the measurement range according to the scene strategy, effectively solving the monitoring delay and judgment error caused by manual range switching in the prior art, and realizing the intelligent operation of the detection equipment in complex fire environment. The system can accurately grasp the best time for range switching by continuously analyzing the characteristic parameters of high-temperature area combined with the preset threshold condition, avoiding measurement saturation or precision loss caused by improper range.

[0026] 2. Due to the adoption of three-dimensional scene modeling and heat source identification mechanism based on multi-device historical data, the system can build a complete heat source distribution map of the monitoring area, and realize accurate positioning and feature analysis of new heat sources, effectively solving the problems of single-device monitoring angle limitation and insufficient heat source identification accuracy in the prior art, and realizing global perception and accurate judgment of fire situation. The system establishes a three-dimensional scene model containing heat source position and characteristics by integrating the historical data of multiple related detection instruments, providing rich reference for heat source identification.

[0027] 3. Due to the adoption of high-temperature area standardized identification and attribute analysis method based on temperature reference value, the system can establish a complete high-temperature area feature description system, and realize accurate division and attribute statistics of the area, effectively solving the problems of high-temperature area boundary ambiguity and incomplete feature extraction in the prior art, and realizing accurate identification and feature quantization of high-temperature area. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 is a flowchart of the temperature range dynamic switching method of the detection instrument based on environmental adaptation in the embodiments of the present application;

[0029] Figure 2 is another flowchart of the temperature range dynamic switching method of the detection instrument based on environmental adaptation in the embodiments of the present application;

[0030] Figure 3 is a schematic diagram of an entity device structure of the detection instrument control system in the embodiments of the present application. DETAILED DESCRIPTION

[0031] The terminology used in the following embodiments of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used in the description of the application, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It also will be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0032] Hereinafter, the terms "first", "second" are used only for the purpose of description, and cannot be understood as implying or suggesting relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, the meaning of "a plurality of" is two or more, unless otherwise specified.

[0033] For the convenience of understanding, the terms involved in the present application are introduced as follows.

[0034] Infrared thermal imaging data is not just traditional image information, it is essentially a fine, digitized temperature distribution matrix. When the infrared thermal imager works, its internal detector array will receive the infrared radiation emitted by the objects in the scene. The system will process and calibrate these radiation intensity signals through complex algorithms, and convert them into a two-dimensional array, where each element (i.e. pixel) value accurately corresponds to the temperature of a specific point in the scene. The color image we usually see on the screen is actually the visualization result of this temperature matrix after "pseudo-color" rendering.

[0035] The system maps different temperature ranges to different colors through a preset color palette (for example, using blue to represent low temperature, and using yellow, red to white to represent gradually increasing temperature), thereby converting the invisible temperature field into visual information that the human eye can intuitively understand. Therefore, each frame of infrared thermal imaging data not only provides a qualitative "where is hot, where is cold" image, but more importantly, provides quantified full-scene temperature data that can be used for accurate analysis.

[0036] The range is the working range set by the temperature measuring device to ensure the measurement accuracy and effectiveness, and the selection behind it is the trade-off between sensitivity and upper limit of measurement.

[0037] Low range (e.g. -20°C to 150°C) is designed for high sensitivity scenarios. In this mode, the entire dynamic range of the detector is concentrated to represent a relatively narrow temperature interval. This means that even very small temperature differences (e.g. tenths of a degree Celsius) can produce significant changes in the digital signal, which can be clearly distinguished on the image. This is crucial for searching for trapped people with a body temperature of about 37°C in a normal temperature environment. However, the cost is a lower upper measurement limit. Once there are objects in the field of view that exceed 150°C (such as flames), the signal output of the corresponding area of the detector will reach saturation, i.e. the maximum output. At this time, a pure white highlight area (i.e. "highlight blur") without details will be displayed on the screen, and the system cannot distinguish whether this area is 200°C or 1000°C, and all the details of the high temperature area are completely lost.

[0038] High range (e.g. 0°C to 1200°C) is used to measure extreme high temperatures. It extends the dynamic range of the detector to a very wide temperature interval, so that the temperature of the target such as the core of the flame can be accurately measured, avoiding signal saturation. However, the sensitivity or "temperature resolution" will decrease accordingly. The signal step that represents a 0.1°C change in low range may represent a 5°C or even higher temperature change in high range. This makes it difficult to find heat sources such as the human body, which have little temperature difference with the environment, in the high range mode, because their heat signals are easily overwhelmed by measurement noise.

[0039] Therefore, dynamic adjustment is the key to effective detection. Rescue personnel need low range high sensitivity to find signs of life when entering unknown buildings; when approaching the core of the fire, the system must intelligently and automatically switch to high range to accurately assess the intensity of the fire, locate the core of the fire and avoid the risk of structural collapse, so as to ensure that rescue personnel can obtain continuous, accurate and blind-free environmental perception during the entire task.

[0040] The method provided by the embodiment is described in the following flow. Please refer to Figure 1 , a flowchart of the method for dynamically switching the temperature range of the detector based on environmental self-adaptation in the embodiment of the present application.

[0041] S101, acquire infrared thermal imaging data, identify high temperature pixel points in the infrared thermal imaging data that exceed a preset temperature threshold as a target high temperature area.

[0042] In the formula, the infrared thermal imaging data represent a scene temperature distribution image collected by an infrared detector array, and contain a correspondence between temperature values and spatial positions; the preset temperature threshold refers to a temperature determination reference value preset according to different scenes, and is used to distinguish a normal temperature region and an abnormally high temperature region, for example, can be set to 40°C in a personnel search and rescue scene, and can be set to 100°C in a fire source positioning scene; the high temperature pixel point represents a minimum image unit in the image whose temperature value exceeds the preset threshold; and the target high temperature region refers to a continuous region formed by a plurality of adjacent high temperature pixel points, and is used to represent an abnormally high temperature region that needs to be monitored.

[0043] After receiving the real-time image data stream of the infrared detector, the detector control system needs to quickly identify the abnormally high temperature region in the scene. Specifically, the detector control system first pre-processes the image data, including non-uniformity correction and temperature calibration, to ensure that the pixel gray value accurately corresponds to the actual temperature value. Then, the detector control system compares the temperature value with the preset threshold pixel by pixel, and marks the pixels exceeding the threshold as high temperature points. Next, the detector control system performs connected component analysis on these high temperature points, and merges the high temperature points adjacent in spatial position into a complete target region. In this process, the detector control system also filters noise regions with too small areas, and performs smoothing processing on the region boundaries, to improve the integrity and accuracy of the target region.

[0044] In some embodiments, the accurate identification and feature extraction of the high temperature region can be achieved in various ways: optionally, the detector control system can use an adaptive threshold method to dynamically adjust the determination threshold according to the overall temperature distribution of the image, first calculate the temperature histogram of the image to analyze the temperature distribution characteristics, then determine the optimal segmentation threshold based on the OTSU algorithm (Otsu method), and finally use morphological operation to optimize the region boundary; optionally, the detector control system can use a multi-scale analysis method, decompose the image into multiple layers through a Gaussian pyramid, detect the high temperature region at different scales, and then fuse the detection results at different scales to achieve accurate identification of high temperature targets of different sizes. It can be understood that other ways such as a target detection method based on deep learning or an edge detection method based on temperature gradient can also be used to identify the high temperature region, which is not limited here. In addition, when the image quality is poor or there is serious occlusion, the target tracking and prediction can be combined with the time sequence information and spatial correlation to improve the robustness of the identification.

[0045] In practical applications, due to the complex and changeable fire environment, multiple high-temperature regions may interact or have sharp temperature fluctuations, leading to inaccurate identification of target regions. To solve this problem, the detector control system adopts a space-time joint analysis strategy: in the spatial dimension, the temperature gradient field is calculated to identify the thermal source boundary, and the boundary is corrected based on the heat conduction effect; in the time dimension, the position and shape of the target region are tracked and predicted using Kalman filtering to suppress the influence of short-term fluctuations. At the same time, the system also establishes a region feature library to store typical feature patterns of different types of heat sources, which can be used to assist in judging the reliability of the target region. For example, for a stable burning fire source, its temperature distribution usually presents a characteristic pattern of high center and decreasing edge, which can be used to distinguish false high-temperature regions.

[0046] S102, calculate the area proportion of the target high-temperature region in the full picture of the infrared thermal imaging data to obtain a high-temperature region proportion value.

[0047] Among them, the area proportion represents the numerical ratio of the total number of pixels of the target high-temperature region to the total number of pixels of the image, which is used to quantify the spatial proportion of the high-temperature region in the field of view; the full picture refers to the current complete field of view range of the infrared thermal imager, which includes all visible monitoring regions; the high-temperature region proportion value represents the normalized area proportion value, which is used to represent the spatial distribution degree of the high-temperature state, for example, when the proportion value is 0.3, it means that 30% of the field of view region is in a high-temperature state.

[0048] After the detector control system completes the identification of the high-temperature region, it needs to evaluate the spatial distribution range of the high-temperature state. Specifically, the detector control system first counts the number of effective pixels in the target high-temperature region, and eliminates isolated pixel points caused by image noise or edge effects. Then, the detector control system obtains the total number of effective pixels of the current field of view, excluding invalid regions caused by lens distortion or detector dead spots. Next, the detector control system obtains the accurate area proportion value through division operation, and performs decimal point calibration and numerical normalization processing.

[0049] When calculating the high-temperature region proportion, the problem of distorted proportion calculation due to field of view obstruction or partial region reflection may occur. To solve this problem, the detector control system adopts a multiple verification mechanism: first, a field of view visibility map is established to mark the regions that may be affected by obstruction or reflection; then, according to the heat conduction model, the temperature value reliability of these regions is evaluated; finally, the unreliable regions are compensated in the proportion calculation. For example, when a strong reflection region is detected, the system estimates the actual temperature value of the region based on the temperature distribution characteristics of the surrounding regions, and corrects the proportion calculation result accordingly. At the same time, the system also records the uncertainty of the correction process to evaluate the reliability of the calculation result.

[0050] In addition, to ensure the comparability of the proportion value under different device states, the system must normalize the changes caused by operations (such as digital zoom) or settings (such as resolution switching). For example, when the user uses the digital zoom function to enlarge a certain local high-temperature area, the number of pixels occupied by this area on the screen will increase significantly. If the proportion is directly calculated, it will be concluded that the high-temperature area has expanded sharply. To solve this problem, the control system will track the current zoom ratio or field of view (FOV) parameter in real time. It will scale the currently calculated pixel proportion value according to the zoom ratio to convert the proportion of the high-temperature area in the unzoomed, standard wide-angle field of view. Similarly, if the resolution of the device is switched from 640x480 to 320x240, the total number of pixels becomes one-fourth of the original. The system will dynamically adjust the area proportion threshold according to the resolution parameter, or convert the calculated pixel number into an equivalent physical area (such as square meters), so that whether it is to observe details or switch to a low-power mode, the area proportion value judgment standard always remains consistent and accurate.

[0051] S103, determine the area proportion threshold and duration threshold corresponding to the current scene mode in the mode strategy library.

[0052] Among them, the mode strategy library refers to a database that stores monitoring parameter configurations under different scenes, which is used to realize the scene adaptation of detection strategy; the scene mode represents a specific fire environment type and its corresponding monitoring requirements, which can be divided into different modes such as initial fire, full combustion, and residual fire monitoring; the area proportion threshold is the area proportion determination standard for triggering the next monitoring action; the duration threshold represents the minimum time period that needs to be observed, which is used to filter the influence of transient fluctuations.

[0053] The detection instrument control system needs to select the appropriate monitoring strategy according to the current fire situation. Specifically, the detection instrument control system first analyzes the current scene characteristics, including temperature distribution mode, heat source quantity, spatial distribution, etc. to determine the type of the current scene. Then, the detection instrument control system accesses the mode strategy library to extract the standard parameter configuration under this scene. Next, the detection instrument control system fine-tunes the standard parameters in combination with real-time environmental factors (such as environmental temperature, humidity, air flow, etc.) to generate the final judgment threshold.

[0054] Note that the determination of the pattern strategy library is a preset process based on professional knowledge and massive data analysis. It is co-constructed by fire experts, thermal imaging engineers, and data scientists. By analyzing a large amount of infrared data collected in real fire cases, fire drills, and laboratory simulations, experts define the corresponding optimal response strategies for different types of fire environment. For example, for the "initial fire" scenario, they may find through data statistics that when the high-temperature area ratio exceeds 5% and lasts for more than 2 seconds, there is a high probability that the fire will spread rapidly. Therefore, "area ratio threshold = 0.05" and "duration threshold = 2s" are stored as a strategy in the library. This database is essentially a complex rule set or lookup table that accurately maps various predictable scene patterns to a set of optimized threshold parameters (including area, time, temperature change rate, etc.), which is the core basis for the detector to achieve "environmentally adaptive" intelligent decision-making.

[0055] Scene patterns are a classification of typical environments that firefighters may encounter, with distinct thermodynamic characteristics and operational objectives. The division of these patterns aims to match the device's behavior to the current task requirements. For example, the following patterns can be defined:

[0056] 1) "Personnel search and rescue mode": suitable for environments with low temperature, where the primary goal is to find vital signs, at which time the system will prefer to use high sensitivity and low range;

[0057] 2) "Fire point investigation mode": suitable for scenarios where small-scale fire points have been discovered and need to be evaluated for their development trend, at which time the threshold settings will be very sensitive to trigger range switching or alarms as soon as possible;

[0058] 3) "Comprehensive burning mode": suitable for large-scale intense burning fire scenes, aiming to analyze the flame structure and find weak points in the fire, at which time the system will work in high range and focus on monitoring the rapid temperature change;

[0059] 4) "Afterburning cleaning mode": suitable for scenarios where the main fire has been extinguished and hidden smoldering fire points need to be found in walls and ceilings, at which time the system needs to return to high sensitivity mode again.

[0060] The device's judgment of the current scene pattern can be manually selected by the user or achieved through a more advanced automatic recognition algorithm. This algorithm will analyze the temperature histogram of the image, the number and shape of high-temperature areas, the rate of temperature change, and other macro features in real time, automatically classify the current environment into the most suitable preset pattern, and trigger the corresponding strategy to achieve fully automatic intelligent detection.

[0061] In the determination of threshold parameters, there may be complex situations where scene features are not obvious or multiple scenes are superimposed, leading to difficulties in policy selection. To solve this problem, the detector control system adopts a multi-mode fusion strategy: first, a scene feature vector is constructed, including temperature distribution, time evolution and other multi-dimensional features; then the similarity between the current scene and each standard mode is calculated to obtain the mode matching weight; finally, the comprehensive threshold parameter is generated by weighted fusion. For example, when both large-area low-temperature anomalies and local high-temperature phenomena are detected, the system will refer to the threshold configurations of multiple related scene modes at the same time to determine the optimal monitoring parameters comprehensively. At the same time, the system will continuously track the evolution of the scene and adjust the mode weight in a timely manner to ensure the dynamic adaptability of the monitoring strategy.

[0062] S104, when the high-temperature area proportion value exceeds the area proportion threshold value, start the timer and determine the high-temperature duration based on the timer.

[0063] Wherein, the high-temperature area proportion value represents the relative area size of the high-temperature area in the current monitoring picture; the area proportion threshold value is a determination criterion for triggering timing, used to determine whether the high-temperature state needs to be further observed; the timer represents a time counting module for recording the high-temperature duration; the high-temperature duration is the cumulative length of time when the high-temperature area proportion value continuously exceeds the threshold value, used to evaluate the stability of the high-temperature state.

[0064] The detector control system needs to monitor and evaluate the persistence of the high-temperature state. Specifically, the detector control system first compares the real-time calculated high-temperature area proportion value with the pre-set area proportion threshold value. When the proportion value first exceeds the threshold value, the detector control system immediately starts the timer and records the starting time stamp. During the timing process, the detector control system continuously monitors the change trend of the proportion value, and if the proportion value decreases below the threshold value, the timer is reset. When the proportion value exceeds the threshold value again, the timing starts again. The detector control system obtains the accurate high-temperature duration by accumulating the effective timing period. At the same time, the system records the proportion value fluctuation during the period, which is used to evaluate the stability of the high-temperature state.

[0065] In some embodiments, the accurate calculation of high-temperature duration can be achieved in various ways: optionally, the detector control system can adopt a segmented timing strategy, first segmenting the high-temperature state by time, then analyzing the change characteristics of the proportion value in each time segment, setting different time weights, and finally obtaining the equivalent duration by weighted accumulation to realize differentiated evaluation of different degrees of high-temperature state; optionally, the detector control system can adopt a fuzzy timing method, by establishing a mapping relationship between the proportion value and the time gain, using a continuous time accumulation model in the threshold interval to avoid timing interruption caused by slight fluctuations in the proportion value. It can be understood that other ways such as state machine-based timing control method or probability statistics-based time estimation method can also be used to realize duration calculation, which is not limited here. In addition, the timing accuracy and sampling frequency can be adjusted according to actual needs to ensure the accuracy of time measurement.

[0066] During timing, the proportion value may fluctuate sharply due to device jitter or environmental interference, affecting the accurate calculation of the duration. To solve this problem, the detector control system adopts a robust timing mechanism: first, the proportion value is filtered by moving average to suppress short-term fluctuations; then, a hysteresis interval is set, and only when the proportion value continuously exceeds or falls below the hysteresis interval will the state of the timer be switched; finally, the state duration is verified to filter out state switching with too short duration. For example, when detecting that the proportion value fluctuates rapidly around the threshold, the system will dynamically adjust the width of the hysteresis interval according to the fluctuation amplitude and frequency characteristics to ensure the continuity and reliability of timing. At the same time, the system will record abnormal fluctuation events for subsequent parameter optimization.

[0067] S105、In the case where the high-temperature duration exceeds the duration threshold, the temperature change rate of the target high-temperature region is calculated according to the continuous multiple frames of infrared thermal imaging data in the high-temperature duration.

[0068] Wherein, the continuous multiple frames represent a series of infrared thermal imaging data collected at fixed time intervals, used to construct a temperature change sequence; the temperature change rate refers to the change amplitude of the average temperature of the target high-temperature region per unit time, used to represent the speed of temperature rise or fall; the high-temperature duration represents the effective time window of the system for continuous monitoring.

[0069] The detector control system needs to evaluate the temperature dynamic change characteristics of the high-temperature area. Specifically, the detector control system first extracts continuous thermal imaging data frames within the high-temperature duration window according to the preset sampling frequency. Then, the detector control system performs temperature statistics on the target high-temperature area in each frame of data, calculates the area average temperature, the maximum temperature, and the temperature distribution characteristics. Next, based on the time series temperature data, the temperature change amount at adjacent time points is calculated by difference calculation. Due to the nonlinear characteristics of temperature change, the system uses piecewise linear fitting method to calculate the temperature change rate at different time scales.

[0070] At the same time, the system also analyzes the acceleration characteristics of temperature change to evaluate the stability of temperature change trend. Specifically, the acceleration of temperature change (i.e. the speed of change) is a more forward-looking dangerous indicator. For example, a temperature rises from 200℃ at a rate of 10℃ / s, and another also rises from 200℃ at a rate of 10℃ / s, but its acceleration is positive (e.g. +2℃ / s 2 ). The danger is completely different. The latter means that the fire is developing out of control and accelerating, the temperature rising rate in the next second will be 12℃ / s, and in the next second it will be 14℃ / s, which may be a precursor to extreme dangerous phenomena such as flashover. The system calculates the acceleration value of temperature change by again performing difference calculation on the calculated continuous multiple temperature change rate values. If the acceleration is continuously positive and exceeds the preset threshold, the system will determine that the trend is "unstable" or "accelerating deterioration", and may trigger the highest level of warning, even if the current temperature and change rate have not reached the threshold for triggering range switching, valuable prediction and decision-making time can be gained for the firefighters.

[0071] When calculating the temperature change rate, the problem of unstable temperature measurement due to complex fire environment may be encountered. To solve this problem, the detector control system adopts multiple verification mechanisms: first, the original temperature data is preprocessed, including denoising, calibration and spatial mean filtering; then, a temperature change prediction model is established to evaluate the rationality of the measurement value in real time; finally, based on the correlation analysis of multiple frames of data, abnormal temperature fluctuations are identified and corrected. For example, when a temperature value that deviates seriously from the change trend is detected in a frame of data, the system will combine the data of adjacent frames for interpolation correction to ensure the accuracy of the change rate calculation. At the same time, the system also records the uncertainty of data correction for evaluating the credibility of the calculation result.

[0072] S106、When the temperature change rate exceeds the preset change threshold, an instruction is sent to the control circuit of the thermal imaging module, so that the target detector switches to the target measurement range to monitor the target high-temperature area.

[0073] The preset change threshold refers to the change rate determination standard for triggering range switching. The thermal imaging module represents the core component unit including the detector, optical system, and control circuit. The control circuit refers to the electronic control unit responsible for range switching and parameter adjustment. The target measurement range represents the new range position suitable for the current temperature range, used to ensure temperature measurement accuracy. The target detector refers to the specific device unit that needs to perform range switching.

[0074] The detector control system needs to achieve intelligent switching of the range based on temperature change characteristics. Specifically, the detector control system first compares the calculated temperature change rate with the preset change threshold. When the change rate exceeds the threshold, the system further analyzes the temperature change trend and predicts the temperature change range in the short term. Then, the detector control system selects the most suitable target range from the available range based on the prediction result. During the selection process, the system determines the hysteresis effect of range switching and the measurement dead zone to ensure the smoothness of the switching process. Next, the system generates control instructions containing target range parameters and sends them to the control circuit of the thermal imaging module through the communication interface. After receiving the instructions, the control circuit adjusts the gain setting and response curve of the detector to complete the range switching.

[0075] At the same time, the system monitors the execution status of the switching process to ensure that the instructions are correctly responded. When the CPU issues a range switching instruction, it does not blindly consider the task completed, but starts an internal timer and waits for the response of the thermal imaging module control circuit. After successfully switching the hardware gain and parameter configuration, the module sends an "switching success" confirmation signal (ACK) to the CPU. If the CPU receives this confirmation signal within the preset timeout time (e.g. 500 milliseconds), it updates the system state and confirms that the device is in the new range. If the confirmation is still not received within the timeout, the system determines that the switching fails. At this time, it starts the fault handling program: first, it may try to resend the instruction several times; if it still fails, the system must immediately alert the user, for example, displaying the "range switching failure" icon on the screen, accompanied by a unique sound and light prompt. This can prevent firefighters from continuing to approach the extreme high temperature area under the assumption that the device has switched to the high range, thereby avoiding misjudgment and danger due to incorrect device state cognition.

[0076] When performing range switching, the problem of frequent range switching due to drastic temperature fluctuations may occur, affecting the stability of measurement. To solve this problem, the detector control system adopts an intelligent anti-shake strategy: first, a state cache mechanism for range switching is established to record the switching history in the short term; then, the switching frequency and temperature change pattern are analyzed to identify possible oscillation; finally, by setting the minimum switching interval and switching condition retention time, unnecessary frequent switching is suppressed. For example, when the system detects that the temperature is fluctuating near the critical value of the two ranges, it will expand the range overlap interval and increase the switching hysteresis time to ensure the stability of the range switching. At the same time, the system will also adaptively adjust the anti-shake parameters according to the actual application scenario, balancing the requirements of response speed and stability.

[0077] The method provided by the embodiment is further described in a more specific flow. Please refer to Figure 2 , another flowchart of the method for dynamically switching the temperature range of the detector based on environmental adaptation in the embodiment of the present application.

[0078] S201, based on the historical posture coordinate data and historical infrared data collected by multiple associated detectors, a three-dimensional scene map containing multiple known heat source positions and known heat source characteristics is constructed.

[0079] Among them, the associated detector means multiple infrared thermal imaging devices working together, which can be other detectors with the same device function located in the same monitoring area as the target detector, or other infrared thermal imaging devices, or historical infrared thermal imaging systems; historical posture coordinate data refers to time series data recording the spatial position and orientation of the device; historical infrared data represents the previously collected temperature distribution image sequence; known heat source position refers to the confirmed heat source spatial coordinates; known heat source characteristics refer to feature descriptions including temperature distribution, shape, etc.

[0080] The detector control system needs to construct a complete monitoring scene model based on historical data. Specifically, the detector control system first time-aligns and space-registers the historical data of multiple detectors to establish a unified coordinate reference system. Then, through a multi-view fusion algorithm, the heat source information collected by different devices is projected into a three-dimensional space to form a preliminary spatial distribution map. Next, the system extracts and classifies heat source features, including temperature distribution patterns, geometric features, and time-varying characteristics. Finally, through spatial interpolation and feature clustering, a three-dimensional scene map containing complete heat source information is constructed.

[0081] During this process, the system determines factors such as device positioning errors and field-of-view overlaps to ensure the accuracy of the scene reconstruction. The pose coordinate data of each detector (provided by sensors such as the built-in inertial measurement unit, IMU) has accumulated errors, i.e., "drift". If the data of multiple devices with errors are directly spliced, ghosting, misplacement, and deformation will occur in the three-dimensional map. The system uses the overlapping areas of the fields of view between multiple detectors to solve this problem. It automatically identifies and matches common, prominent heat source feature points or structural feature points in these overlapping areas. Then, the system runs a global optimization algorithm that simultaneously fine-tunes the pose coordinates of all detectors and the spatial positions of three-dimensional feature points, with the goal of minimizing the position error of all matched feature points "projected" back in their respective images from the actual observed positions. In this way, the system "locks" all device data into a unified and high-precision coordinate system, greatly eliminating the effects of single-device positioning errors, and thus generating a clear, consistent, and ghost-free global three-dimensional scene map.

[0082] In some embodiments, the construction of the three-dimensional scene map can be achieved in various ways: optionally, the detector control system can use a dynamic reconstruction method based on SLAM (simultaneous localization and mapping), first establish the position association between devices through feature matching, then optimize the spatial position estimation using the Bundle Adjustment (BA) method, and finally realize real-time updating of the scene by fusing heat source information through a probabilistic graph model; optionally, the detector control system can use a hierarchical modeling strategy, decompose the scene into two levels of static background and dynamic heat sources, and model and update them respectively to improve the reconstruction efficiency. It can be understood that other ways such as scene understanding methods based on deep learning or spatial division methods based on voxels can also be used to realize scene reconstruction, which are not limited here.

[0083] In addition, the system should have the functions of compressed storage and fast retrieval of scene models. The original three-dimensional point cloud data (including the XYZ coordinates, temperature, timestamp, etc. of each point) is extremely large in volume, and direct storage will quickly consume the storage space of the device, and the efficiency is low when querying. Therefore, the system uses an efficient spatial data structure such as "octree". The octree recursively divides the entire three-dimensional space into eight subcubes until the point cloud density in each subcube is below a certain threshold or reaches the minimum size. This structure naturally compresses the data, as large areas of empty space without point cloud are represented at higher levels of the tree and do not need further subdivision. At the same time, it is a powerful spatial index. When a specific area needs to be retrieved (for example, "show the heat sources within 5 meters around me"), the system only needs to traverse the relevant branches of the octree, and can quickly eliminate most of the irrelevant spatial data, with a query speed several orders of magnitude faster than point-by-point scanning, thereby ensuring the real-time interaction and analysis capabilities of the three-dimensional map.

[0084] When building a three-dimensional scene map, data incompleteness problems may occur due to changes in device position or scene occlusion. To solve this problem, the detector control system uses a robust reconstruction strategy: first, establish a data integrity evaluation mechanism to identify data gaps and uncertain areas in the scene; then, based on physical models and prior knowledge, supplement the data, such as using a heat conduction model to estimate the temperature distribution in the occluded area; finally, through temporal fusion of multi-time data, improve the completeness of scene reconstruction. For example, when a certain area cannot be directly observed temporarily, the system will combine historical data and observation results from surrounding areas to infer the possible state of the area and mark the credibility of the reconstruction results.

[0085] In some embodiments, the detector control system will build a heat source spatial distribution model based on historical data, that is, the detector control system will read the historical pose coordinate data and historical infrared data collected by multiple associated detectors from the state record database, and construct a data index table according to the collection device ID and timestamp; according to the timestamp order in the data index table, the historical pose coordinate data is converted and registered in the coordinate system to generate spatial point cloud data in a unified coordinate system with time sequence correlation; the historical infrared data is mapped to the corresponding spatial point cloud position in the spatial point cloud data, and the heat source feature parameters of multiple spatial points in the spatial point cloud data are calculated; according to the spatial point cloud data and the heat source feature parameters, a three-dimensional scene map containing known heat source positions and known heat source features is constructed.

[0086] wherein the state record database represents a central storage system storing multi-device historical data; the historical pose coordinate data refers to time-series data recording the spatial position and orientation of devices; the historical infrared data represents temperature distribution images collected previously; the data index table refers to a data retrieval structure established based on device identification and time information; the coordinate system conversion represents spatial mapping between different device coordinate systems; the registration refers to aligning spatial data at different time to a unified reference system; the spatial point cloud data represents a discrete set of sampling points describing a three-dimensional scene; the heat source feature parameter refers to a numerical index describing the temperature, shape, and other characteristics of a heat source; and the three-dimensional scene map is used to represent a complete monitoring space model.

[0087] When a complete monitoring space model needs to be constructed, the detector control system needs to perform fusion processing on multi-source historical data. Specifically, the detector control system first retrieves historical data of multiple detectors from the database and establishes a two-dimensional index structure based on device ID and time stamp. Then, the system processes the pose data in chronological order, converts the coordinate data of each device to a unified world coordinate system through rigid body transformation and calibration parameters. Next, the system performs spatio-temporal registration on the converted data to eliminate errors caused by device drift and time asynchronization. On this basis, the system maps the infrared image data to the corresponding spatial position and fills the temperature information of the spatial point cloud through an interpolation algorithm. Finally, the system extracts the heat source features of each spatial point, including temperature value, gradient, shape feature, etc., and constructs a three-dimensional scene model with complete heat source information.

[0088] In some embodiments, the detector control system performs regional heat source activity analysis and monitoring strategy optimization, i.e., the detector control system groups the historical infrared data of each collection device according to the monitoring region based on the data index table, calculates the spatial distribution density and temperature cumulative duration of infrared heat sources in each monitoring region, identifies the heat source gathering points and heat diffusion channels in the region according to the spatial distribution density and temperature cumulative duration, constructs a heat source activity map, superimposes the heat source activity map on the corresponding pose coordinate data in the data index table to establish a mapping relationship between the heat source activity region and the spatial position, and based on the mapping relationship, classifies the spatial regions in the three-dimensional scene map according to the heat source activity level and adjusts the sampling density and monitoring frequency of the target detector in different regions according to the classification results.

[0089] The monitoring area represents an independent observation unit divided in space; the spatial distribution density refers to the number of heat sources in a unit of space; the temperature accumulation time length represents a time measure of the existence of a heat source; the heat source gathering point refers to a spatial position with a higher heat source density; the heat diffusion channel represents the main path of heat transfer; the heat source activity map refers to a visual model describing the dynamic distribution of heat sources; the spatial position mapping relationship represents the correspondence between heat source activity and actual position; the heat source activity level is used to represent the intensity of heat source activity in different regions; the sampling density refers to the spatial resolution of data collection; and the monitoring frequency represents the time interval of data update.

[0090] When the detector control system needs to optimize the configuration of monitoring resources, the spatio-temporal distribution characteristics of heat sources need to be analyzed. Specifically, the detector control system first groups historical data in space based on the data index table to determine the data set of each monitoring area. Then, the system calculates the spatial density distribution of heat sources in each area, including the number, distribution range and aggregation degree of heat sources, and simultaneously counts the duration characteristics of heat sources. Next, the system identifies heat source dense areas through a density clustering algorithm and determines heat propagation paths using temperature gradient analysis to construct a dynamic distribution map reflecting the activity rules of heat sources. The system maps these activity characteristics to actual spatial positions and determines the heat source activity level of each area through multi-level analysis.

[0091] Finally, based on the activity level classification results, the system dynamically adjusts the data collection strategy in each area, increases the sampling density and monitoring frequency in key areas, and realizes the optimal configuration of monitoring resources. Specifically, this dynamic adjustment is an intelligent scheduling mechanism for the internal computing resources of the detector control system. For example, for an area rated as “high activity level” (such as a heat diffusion channel or a heat source with a sharp temperature rise that has been identified), the system will assign it the highest priority. This means that for each frame of infrared data entering this area, the system will invest all computing resources for the most detailed analysis, including high-precision calculation of its temperature change rate, acceleration and morphological evolution. This is equivalent to increasing the “monitoring frequency” to the upper limit of the physical frame rate of the device. Conversely, for an area rated as “low activity level” (such as a background heat source with stable temperature and no change for a long time), the system will reduce its “monitoring frequency”. It may not analyze it frame by frame, but only perform a complete feature update every 10 frames or longer. In the middle frames, the system may only do a simple existence check. In this way, the system frees up valuable processor cycles from non-critical areas and concentrates them on monitoring and predicting the most dangerous and rapidly changing parts of the fire field, thereby maximizing the response speed and analysis depth of critical threats under limited on-board computing power, achieving on-demand allocation and efficient use of monitoring resources.

[0092] In some embodiments, the heat source activity analysis and monitoring optimization can be achieved in various ways: optionally, the detector control system can use a spatiotemporal clustering method, first construct a spatiotemporal density field of the heat source, then use density peak detection to identify the activity center, and finally determine the activity pattern through trajectory analysis to accurately describe the behavior of the heat source; optionally, the detector control system can use an adaptive grid division strategy, dynamically adjust the size and shape of the monitoring grid, optimize the sampling layout, and improve the monitoring efficiency. It can be understood that other ways such as machine learning-based pattern recognition methods or flow field analysis-based diffusion modeling methods can also be used to achieve activity feature extraction, which is not limited here.

[0093] S202, acquire infrared thermal imaging data, and extract a high-temperature combined area of all high-temperature pixel points in the infrared thermal imaging data.

[0094] Wherein, the infrared thermal imaging data represents a temperature distribution image collected at the current moment; the high-temperature pixel point refers to a basic image unit whose temperature value exceeds the determination standard; the high-temperature combined area represents a continuous area set composed of multiple adjacent high-temperature pixel points, used to represent a candidate area where a heat source may exist; the extraction process refers to a data processing procedure for converting original image data into structured heat source area information.

[0095] The detector control system needs to identify all potential heat source areas from real-time images. Specifically, the detector control system first pre-processes the original infrared image, including noise suppression, temperature calibration, and image enhancement. Then, the system uses an adaptive threshold segmentation algorithm to determine the high-temperature determination standard according to the overall temperature distribution characteristics of the image. Next, the pixel points exceeding the threshold are labeled, and adjacent high-temperature pixels are merged into complete combined areas through a region growing algorithm. During the merging process, the system will combine temperature gradients and spatial continuity to ensure the rationality of the region division. Finally, preliminary shape analysis and boundary extraction are performed on each combined area to prepare for subsequent feature matching.

[0096] When extracting the high-temperature combined area, the problem of over-segmentation or under-segmentation of the area due to uneven temperature distribution may occur. To solve this problem, the detector control system uses an adaptive region merging strategy: first, analyze the region characteristics of the preliminary segmentation results, including temperature distribution, shape features, and boundary strength; then establish region merging criteria, determine the merging threshold according to physical constraints and statistical characteristics; finally, adjust the region through iterative optimization. For example, when detecting that two adjacent regions have similar temperature distribution characteristics and weak boundary gradients, the system will merge them into a single combined area. At the same time, the system also records the uncertainty of the merging process for subsequent region verification.

[0097] S203, calculate the estimated position of the combined region in the three-dimensional scene map according to the image features and pose coordinate features of the high-temperature combined region.

[0098] wherein the image features represent the shape, size, temperature distribution, etc. characteristic parameters of the high-temperature combined region in the two-dimensional image; the pose coordinate features refer to the spatial position, orientation, etc. three-dimensional positioning information of the detector; the estimated position represents the spatial coordinates of the heat source obtained by mapping the two-dimensional image information to the three-dimensional space; and the three-dimensional scene map refers to a three-dimensional monitoring environment model containing spatial position relationship, used to support heat source positioning and tracking.

[0099] The detector control system needs to convert the heat source information in the two-dimensional image into a three-dimensional spatial position. Specifically, the detector control system first extracts the image feature parameters of the high-temperature combined region, including geometric features such as region centroid, principal axis direction, area size, and thermal features such as temperature distribution and boundary gradient. Then, combined with the real-time pose data of the detector, a transformation matrix from the image coordinate system to the world coordinate system is established. Next, through the perspective projection model, the two-dimensional feature points are projected into the three-dimensional space, while the lens distortion and projection error are corrected. Finally, based on the spatial distribution of multiple feature points, the spatial position coordinates of the heat source are obtained through least squares estimation.

[0100] In some embodiments, the estimation of the spatial position of the heat source can be realized in various ways: optionally, the detector control system can use a three-dimensional reconstruction method based on feature matching, first track the heat source feature points in the continuous image sequence, then calculate the spatial coordinates through the principle of triangulation, and finally use Kalman filtering to optimize the position estimation, realizing dynamic tracking of the heat source position; optionally, the detector control system can use a depth estimation method, by analyzing the size change and temperature decay characteristics of the heat source, combined with prior knowledge to establish a depth prediction model, to improve the accuracy of position estimation. It can be understood that other ways such as multi-sensor fusion based positioning method or spatial reasoning method based on probability graph model can also be used to realize position estimation, which is not limited here. In addition, the system should have the function of evaluating the uncertainty of position estimation.

[0101] S204, determine multiple known heat source regions and new heat source regions as target high-temperature regions based on the estimated position, known heat source position and known heat source features.

[0102] wherein the estimated position represents the coordinate estimation value of the newly detected heat source in the three-dimensional space; the known heat source position refers to the confirmed heat source spatial coordinates in the historical data; the known heat source features represent the feature description set containing temperature distribution, shape, etc. parameters; the known heat source region refers to the verified historical heat source position; the new heat source region represents the first appearing heat source position; and the target high-temperature region refers to the set of all heat source regions that need to be monitored.

[0103] The detector control system needs to classify and associate the detected heat sources. Specifically, the detector control system first spatially matches the estimated positions with the known heat source positions in the three-dimensional scene map, calculates the distance and feature similarity between each heat source. Then, based on the set matching threshold, it identifies heat source pairs with high similarity and marks them as different time observations of the same heat source. For the estimated positions that cannot be found, the system determines them as new heat sources and establishes new tracking records. Then, the system updates the feature description of each heat source, including position, temperature and shape parameters. Finally, all known heat sources and new heat sources are uniformly managed to form a complete target high-temperature region set.

[0104] In some embodiments, the classification and association of heat sources can be achieved in various ways: optionally, the detector control system can use a multi-target association method based on graph theory, first construct a bipartite graph structure between heat sources, then solve the optimal matching scheme through the Hungarian algorithm, and finally optimize the association result using temporal consistency constraints to achieve stable tracking of multiple heat sources; optionally, the detector control system can use a probabilistic data association method, by establishing a heat source motion model and an observation model, using Bayesian inference to calculate the association probability, and improving the accuracy of heat source identification. It can be understood that other ways such as target tracking methods based on deep learning or classification methods based on multi-feature fusion can also be used to achieve heat source association, which are not limited here. In addition, the system should have the processing capability of heat source splitting and merging.

[0105] In some embodiments, the detector control system will identify the identity and update the status record of the heat source region, that is, the detector control system will obtain the known identity of the known heat source region, and bind a new identity to the new heat source region; bind real-time temperature data, time stamp and collection device ID to known identity and new identity, generate data update record, and upload data update record to status record database.

[0106] Among them, the known heat source region represents the heat source position recorded in the historical data; the known identity refers to the unique code of the historical heat source; the new heat source region represents the first detected heat source position; the new identity refers to the unique code allocated for the new heat source; the real-time temperature data represents the current temperature measurement value; the time stamp refers to the accurate time of data collection; the collection device ID is used to represent the unique identification of different detectors; the data update record represents the structured data containing complete information of the heat source; the status record database refers to the central data warehouse storing the heat source status information.

[0107] After completing the identification and classification of heat source regions, the detector control system needs to establish a complete heat source information management mechanism. Specifically, the detector control system first accesses the state record database to obtain the identity information of known heat sources. Then, the system generates a globally unique identity for each new heat source to ensure that it does not duplicate existing identities. Next, the system binds the real-time collected temperature data, accurate time stamp, and device identification information with the corresponding identity, forming a structured data record. During the binding process, the system verifies the integrity and consistency of the data. Finally, the system updates the records in real time to the state record database to ensure information sharing between multiple devices.

[0108] In some embodiments, the identification management and state update of heat source information can be achieved in various ways: optionally, the detector control system can use a distributed identification management scheme, first generate a temporary identity locally and perform conflict detection, then synchronize with the central database to verify the uniqueness of the identity, and finally confirm the allocation of a permanent identity to achieve consistent management in a multi-device environment; optionally, the detector control system can use an incremental update strategy to achieve efficient synchronization and state tracking of data by establishing a differential data packet and version control mechanism. It can be understood that other ways such as distributed ledger technology based on blockchain or real-time synchronization mechanism based on message queue can also be used to implement information management, which is not limited here.

[0109] In addition, the system also needs to combine the needs of data compression storage and fast retrieval, and can improve efficiency by establishing a multi-level cache and index optimization mechanism. When reading a large amount of historical data from the state record database, efficient retrieval is a prerequisite for building a three-dimensional map. The system not only linearly reads data, but also establishes multi-dimensional indexes when data is warehoused. For example, it uses spatial index structures such as R-tree or k-d tree to organize pose coordinate data, so that it can quickly query "which devices have collected data in a certain three-dimensional space region during a certain time period". At the same time, the system also calculates the hash value of the extracted heat source feature parameters (such as temperature peak value, shape descriptor), and establishes a feature hash table. This allows the system to quickly find similar historical heat sources when performing heat source matching and identification, without the need to traverse the entire database, greatly improving the efficiency of data association and scene construction.

[0110] In the process of heat source information management, there may be a problem of data synchronization not being timely due to network delay or equipment failure. To solve this problem, the detector control system adopts a reliability guarantee mechanism: first, maintain a data cache pool locally to ensure continuous working ability in the case of network disconnection; then, through data checksum and retransmission mechanism, ensure the complete transmission of information; finally, establish a data recovery mechanism to automatically supplement the missing data after communication recovery. For example, when a certain detector temporarily cannot connect to the central database, the system will store the update record in the local cache and record the detailed timing information. After the network is restored, it will synchronize the data in time sequence one by one to ensure the continuity and consistency of the data. At the same time, the system also maintains multiple copies of data backup to prevent information loss.

[0111] S205, calculate the area proportion of the target high-temperature region in the full picture of the infrared thermal imaging data to obtain the high-temperature region proportion value.

[0112] Referring to step S102, the detector control system will calculate the ratio of the number of pixel points of the target high-temperature region to the total number of pixel points.

[0113] S206, determine the area proportion threshold value and the duration threshold value corresponding to the current scene mode in the mode strategy library.

[0114] Referring to step S103, the detector control system will select appropriate threshold parameters from the strategy library according to the current environmental characteristics.

[0115] S207, when the high-temperature region proportion value exceeds the area proportion threshold value, start the timer and determine the high-temperature duration based on the timer.

[0116] Referring to step S104, the detector control system will start the timing module to record the duration of the high-temperature state.

[0117] S208, calculate the temperature average value of all pixel points in the current field of view of the detector as the temperature reference value.

[0118] Wherein, the field of view represents the current observable image range of the detector; the pixel point refers to the smallest temperature sampling unit in the image; the temperature average value represents the arithmetic average result of the temperature of all effective pixel points in the field of view; the temperature reference value refers to the reference temperature for subsequent high-temperature judgment, which is used for adaptive identification of abnormal temperature regions.

[0119] The detection device control system needs to establish a temperature baseline for the current scene. Specifically, the system first verifies the validity of the temperature data within the field of view, eliminating invalid pixels caused by equipment malfunction or environmental interference. Then, the system performs statistical analysis on the temperature values ​​of the valid pixels, including calculating a temperature distribution histogram and basic statistical characteristics. Next, it calculates the overall temperature average using a weighted average method, where the weights can be adjusted according to the reliability and importance of the pixels. Finally, the system uses the calculated average value as a temperature reference value for subsequent high-temperature point identification.

[0120] In some embodiments, the temperature reference value can be calculated in several ways: Optionally, the detector control system can employ an adaptive zonal statistical method, first dividing the field of view into multiple sub-regions, then calculating the local mean of each region, and finally obtaining the global reference value through weighted summation of regional features, thereby effectively handling the unevenness of temperature distribution; Optionally, the detector control system can employ a temporal filtering method, reducing the impact of random fluctuations and improving the stability of the reference value by performing a moving average on multiple consecutive frames of data. It is understood that other methods, such as cluster-based background modeling or statistical feature-based anomaly detection, can also be used to calculate the reference value, and are not limited here.

[0121] S209. Traverse each pixel in the current field of view and mark pixels whose temperature values ​​are higher than a preset multiple of the temperature reference value as high temperature points.

[0122] Here, traversal refers to the process of visiting and processing each pixel in the field of view one by one; temperature value refers to the actual temperature measurement result corresponding to each pixel; preset multiplier represents the temperature ratio threshold for judging high temperature points, which is used to distinguish between normal temperature and abnormal high temperature; high temperature point refers to abnormal pixel point whose temperature is significantly higher than the reference value, which is used to mark the location of potential heat source.

[0123] The detector control system needs to identify abnormally high-temperature points based on temperature reference values. Specifically, the detector control system first establishes a processing queue for pixels to ensure that each pixel is processed in an orderly manner. Then, the system reads the temperature value of each pixel one by one and calculates the ratio with the temperature reference value. Next, the calculated ratio is compared with a preset multiple threshold. When the ratio exceeds the threshold, the pixel is marked as a high-temperature point, and its location coordinates and specific temperature value are recorded. During the processing,

[0124] The system combines the spatial relationship of the pixel points and the temperature continuity to avoid misjudgment caused by local noise. This combination is achieved through a context verification filtering step. After a pixel point is preliminarily marked as a "high-temperature point" because its temperature value exceeds a "preset multiple of the temperature reference value", the system does not immediately confirm it, but checks its surrounding neighborhood (for example, a 3x3 or 5x5 pixel window). If the preliminarily marked high-temperature point has all its neighboring pixels with temperatures much lower than the reference value, it appears as an isolated "spike", and the system determines that it is a false signal caused by detector transient noise or a dead pixel with a high probability, and cancels its "high-temperature point" mark. On the contrary, if the temperature of a pixel point is slightly lower than the marking threshold, but multiple of its neighboring pixels have been marked as high-temperature points, the system can, based on the principle of "temperature continuity", in turn "promote" it to a high-temperature point. This process effectively filters out random "salt and pepper noise" and ensures that only those pixel points that form a certain aggregation in space and conform to the physical law of heat distribution are finally confirmed as valid high-temperature points. Finally, the system generates a marking map containing all high-temperature point information, preparing for subsequent regional analysis.

[0125] In identifying high-temperature points, local misjudgment problems caused by uneven temperature distribution may occur. To solve this problem, the detector control system uses a context-aware strategy: first, analyze the temperature distribution characteristics of the region where the pixel point is located, and establish a local statistical model; then, combine the spatial position relationship to evaluate the reliability of the temperature anomaly; finally, filter out isolated abnormal points through multi-scale verification. For example, when the temperature of a certain pixel point is significantly higher than the reference value but does not match the surrounding temperature distribution, the system will further analyze its spatial context characteristics and determine whether to mark it as a high-temperature point through comprehensive judgment. At the same time, the system also records the confidence of the judgment, which is used for subsequent regional merging analysis.

[0126] S210, performing connected component analysis on the high-temperature points, merging adjacent high-temperature points into standard high-temperature regions, and assigning a unique identifier to each standard high-temperature region.

[0127] Among them, the connected component analysis represents the process of identifying and processing a set of adjacent high-temperature points; adjacent high-temperature points refer to multiple high-temperature pixels that are closely connected in spatial position; standard high-temperature region represents a complete heat source region after standardized processing; unique identifier refers to a code used to distinguish different high-temperature regions, which is used for subsequent tracking and management.

[0128] The detector control system needs to organize the discrete high-temperature points into structured regions. Specifically, the detector control system first establishes a pixel adjacency graph and defines the connectivity rules between high-temperature points. Then, the system uses a depth-first or breadth-first search algorithm to recursively traverse all adjacent high-temperature points from any high-temperature point until the complete region boundary is determined. Next, the system performs shape optimization and boundary smoothing on each identified connected region to remove irregular edges caused by noise. Finally, the system assigns a globally unique identity to each processed standard high-temperature region and establishes a region index table.

[0129] During processing, the system combines the minimum size limit and shape constraint of the region to ensure the rationality of the division result. The "minimum size limit" is a key post-processing step. After the connected component analysis is completed, many high-temperature regions of different sizes are obtained. The system sets a threshold, for example, "5 pixels". Any connected region composed of less than 5 pixels will be directly ignored because they are too small and are likely to be residual noise or physically meaningless hot spots. The "shape constraint" is a more advanced filtering logic that judges whether a region is "reasonable" based on prior knowledge. The system calculates the shape descriptor, such as "aspect ratio" or "compactness", for each region. For example, a real fire source or heat source usually has a relatively compact shape, and it is unlikely to be a perfect diagonal line that spans the entire screen with only a single pixel width. The latter is often a stripe noise generated by the readout circuit failure of the sensor array. Therefore, the system can set rules such as "discard regions with aspect ratio greater than 20:1" to filter out these "false" regions generated by hardware artifacts that do not conform to the shape of natural heat sources, ensuring that the "standard high-temperature regions" in the final output are all potential targets with high confidence for further analysis.

[0130] In some embodiments, the connectivity analysis of high-temperature regions can be implemented in various ways: optionally, the detector control system can use a multi-level merging strategy, first form initial sub-regions based on local connectivity, then design merging criteria according to temperature gradient and shape characteristics, and finally obtain the final standard region through iterative optimization to accurately describe complex heat source shapes; optionally, the detector control system can use an adaptive grid method to adapt to the heat source boundary by constructing an irregular grid structure to improve the accuracy of region division. It can be understood that other ways such as region growing method based on graph segmentation or boundary extraction method based on morphological operation can also be used to implement the connected component analysis, which is not limited here. In addition, the system should have the dynamic ability to handle region splitting and merging.

[0131] When performing connected component analysis, the problem of ambiguous region boundary caused by unobvious temperature gradient may be encountered. In view of this problem, the detector control system adopts a multi-feature fusion strategy: first, a boundary saliency map is constructed by combining temperature value, gradient information and texture features; then, the active contour model is used to accurately position the region boundary; finally, the region is closed and segmented by the boundary tracking algorithm. For example, when there is a temperature transition zone between two heat source regions, the system analyzes the temperature variation trend and spatial structure features, and determines the optimal segmentation position through adaptive threshold.

[0132] S211, record the pixel number, region centroid coordinates and region boundary coordinates of each standard high-temperature region, and construct a region attribute list.

[0133] Among them, the pixel number represents the total number of pixel points contained in the standard high-temperature region, which is used to represent the size of the region; the region centroid coordinates refer to the geometric center position of the high-temperature region; the region boundary coordinates represent the key point set describing the region contour; the region attribute list refers to the structured data record containing multiple feature parameters, which is used to completely describe the characteristics of the high-temperature region.

[0134] The detector control system needs to extract and record the key feature parameters of each standard high-temperature region. Specifically, the detector control system first performs pixel statistics on each region, calculates the area and pixel distribution characteristics of the region. Then, the system calculates the region centroid by weighted average method, where the weight can be set based on the pixel temperature value to highlight the influence of the high-temperature center. Then, the system uses the boundary tracking algorithm to extract the region contour point sequence in the clockwise or counterclockwise direction, and simplifies the key points to obtain a compact boundary description. Finally, the system organizes all the extracted feature parameters into a standardized attribute list, including basic geometric features, statistical features and topological features, etc. multiple dimensions.

[0135] In the feature extraction process, the system will balance the computational efficiency and storage overhead to ensure the completeness and practicality of the feature representation. This balance is reflected in the strategy of feature selection. A "complete" feature set may contain hundreds of dimensions, such as complex texture features (such as gray level co-occurrence matrix) and high-order moments, but this will bring huge real-time computing burden and storage pressure. Therefore, the system will preferentially extract the core features with small amount of calculation but large amount of information. For example, for the boundary of a region, the system will not store the coordinates of all the pixels that constitute the boundary, but may use a polygon approximation algorithm (such as the Douglas-Peucker algorithm) to approximate the contour with a few key vertices, which greatly compresses the storage space while retaining the basic shape information of the region. The system will ensure that basic but crucial features such as area, centroid, maximum / average temperature are calculated efficiently in real time, while more complex features (such as skewness, kurtosis of temperature distribution) may be set to "on-demand calculation" - that is, only when a region is marked as particularly worthy of attention, will these deeper level analysis be triggered.

[0136] In constructing the region attribute list, the problem of inaccurate feature extraction due to complex region shape or irregular temperature distribution may be encountered. To solve this problem, the detector control system adopts an adaptive feature extraction strategy: first, dynamically adjust the accuracy and method of feature extraction according to the complexity of the region; then evaluate the reliability of the extraction results through a feature verification mechanism; finally, establish a feature compensation mechanism to handle abnormal situations. For example, when encountering a complex heat source with multiple sub-regions, the system will adopt a hierarchical feature extraction scheme, first describing the overall features, and then refining the local structure to ensure the accuracy and completeness of the feature description. At the same time, the system also maintains the temporal association of features to support region evolution analysis.

[0137] It should be noted that the region attribute list is not a simple list, but a structured, information-rich "digital file" or "data object" that provides a comprehensive feature portrait for each identified standard high-temperature region. This "file" is the basis for all subsequent advanced analysis (such as tracking, early warning, identification). Its typical content includes at least:

[0138] 1. Unique identifier (ID): a unique code assigned in S210 for intra-frame and cross-frame tracking.

[0139] 2. Geometric properties: total pixel count (i.e. area), geometric centroid image coordinates (x, y), bounding box coordinates, and simplified boundary point sequence.

[0140] 3. Thermodynamic properties: maximum temperature, average temperature, and standard deviation of temperature within the region (used to measure the uniformity of the internal temperature of the region).

[0141] 4. Topology and association attributes: IDs pointing to other regions that are spatially adjacent to it, and a parent ID indicating which region of the previous frame it evolved from in the time series.

[0142] 5. Timestamp: records the time when the region was first detected, and the duration of continuous tracking. By constructing such a detailed attribute list, the system converts the raw, unstructured pixel data into highly structured knowledge units that can be understood and analyzed by machines.

[0143] In some embodiments, the detector control system implements high-temperature region early warning expansion and region merging analysis, that is, the detector control system expands the boundary coordinates of each standard high-temperature region based on the region attribute list, adds a preset distance early warning buffer zone outside the original boundary; adds the pixel points of the early warning buffer zone to the region attribute list, expands the region boundary coordinate record, and obtains an expanded region attribute list; extracts the centroid coordinates of each standard high-temperature region after expansion from the expanded region attribute list, calculates the distance relationship between adjacent regions, and constructs a region distance index; according to the region distance index, identify the region combination whose distance is less than the preset distance threshold, and generate a merging warning information.

[0144] Wherein, the region attribute list represents a data structure recording the characteristic parameters of the high-temperature region; the boundary coordinates refer to the spatial position data describing the region contour; the early warning buffer zone represents the safety monitoring range expanded outside the original boundary; the preset distance refers to the expansion width of the buffer zone; the expanded region attribute list represents the complete feature record containing the buffer zone information; the centroid coordinates refer to the geometric center position of the region; the region distance index represents the data structure of the spatial relationship between adjacent regions; the preset distance threshold is the distance judgment standard for triggering the merging warning; the merging warning information is used to represent the early warning prompt of possible region merging.

[0145] When the detector control system needs to predict the risk of heat source diffusion, it needs to establish a complete early warning mechanism. Specifically, the detector control system first reads the boundary information of each standard high-temperature region, calculates the buffer zone range according to the preset safety distance parameter. Then, the system uses morphological dilation algorithm to expand the original boundary, generates a new early warning boundary, and adds the buffer zone pixels to the attribute list. Next, the system recalculates the geometric features of the expanded region, including updating the centroid position and boundary description. Based on the updated centroid coordinates, the system constructs a spatial adjacency relationship graph, calculates the shortest distance between regions, and establishes a distance index table. Finally, the system identifies the region combination that may merge by comparing the distance value with the preset threshold, and generates warning information containing the risk level and the predicted merging time.

[0146] In the process of constructing the early warning area, the problem of inaccurate buffer zone generation due to irregular shape of the heat source may be encountered. To solve this problem, the detector control system adopts a contour adaptive strategy: first, the characteristics of the area boundary are analyzed to identify the main direction and curvature change of the boundary; then, a variable buffer distance function is designed according to the boundary characteristics to dynamically adjust the expansion range at different positions; finally, the boundary smoothing algorithm is used to optimize the shape of the buffer zone. For example, when the heat source area presents an elongated or irregular shape, the system will adjust the generation parameters of the buffer zone according to the local characteristics of the boundary and the diffusion trend of the heat source, to ensure the rationality of the early warning range. At the same time, the system also records the uncertainty of the buffer zone generation, which is used to evaluate the reliability of the early warning.

[0147] S212, when the high temperature duration exceeds the duration threshold, the temperature change rate of the target high temperature area is calculated according to the infrared thermal imaging data of the continuous multiple frames in the high temperature duration.

[0148] Referring to step S105, the detector control system analyzes the temperature change trend of the target area in the continuous frames.

[0149] S213, when the temperature change rate exceeds the preset change threshold, an instruction is sent to the control circuit of the thermal imaging module to make the target detector switch to the target measurement range to monitor the target high temperature area.

[0150] Referring to step S106, the detector control system sends a range switching instruction to the control circuit.

[0151] In the embodiments of the present application, due to the adoption of the multi-dimensional analysis method based on the high temperature area proportion, duration and temperature change rate, and the systematic solution including the scene strategy library, three-dimensional scene map and early warning buffer zone, the intelligentization and collaboration of temperature monitoring can be realized. At the single device level, the adaptive switching of the range is realized by real-time calculation of multiple parameters and dynamic adjustment of the monitoring strategy combined with the scene characteristics; at the multi-device level, information sharing and collaborative monitoring between devices are realized by establishing a unified three-dimensional scene model and heat source activity map. This innovative technical solution effectively solves the problems of range switching lag, independent devices and insufficient early warning capability in traditional technology, and further realizes a more accurate, reliable and forward-looking temperature monitoring system. This scheme not only improves the measurement accuracy and system reliability, but also enhances the risk prevention and control capability of the system through the early warning mechanism, providing a strong guarantee for industrial safety production.

[0152] The detector control system in the embodiments of the present application will be described from the perspective of hardware processing. Please refer to Figure 3 , which is a schematic diagram of an entity device structure of the detector control system in the embodiments of the present application.

[0153] It should be noted that,Figure 3 The illustrated structure of the detector control system is only one example and should not limit the function and range of use of the embodiments of the present application.

[0154] As shown in Figure 3 the detector control system includes a CPU 301 which can perform various appropriate actions and processes in accordance with a program stored in a ROM 302 or a program loaded into a RAM 303 from a storage section 308, such as performing the methods described in the above embodiments. In the RAM 303, various programs and data required for system operation are also stored. The CPU 301, the ROM 302, and the RAM 303 are connected to each other through a bus 304. An I / O interface 305 is also connected to the bus 304.

[0155] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, a push button switch, and the like; an output section 307 including a Liquid Crystal Display (LCD), an audio output device, an indicator, and the like; the storage section 308 including a hard disk, and the like; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as necessary. A removable media 311 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, and the like is attached to the drive 310 as necessary, so that a computer program read therefrom is installed into the storage section 308 as necessary.

[0156] In particular, the processes described above with reference to the flow charts can be implemented as computer software programs in accordance with the embodiments of the present application. For example, the embodiments of the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing computer programs for executing the methods illustrated in the flow charts. In such embodiments, the computer program can be downloaded and installed from a network via the communication section 309, and / or installed from the removable media 311. When the computer program is executed by the CPU 301, various functions defined in the present application are performed.

[0157] The flow charts and block diagrams in the drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flow charts or block diagrams can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures.

[0158] Specifically, the detector control system of the embodiment includes a processor and a memory, and the memory stores a computer program. When the computer program is executed by the processor, the computer program implements the environment-adaptive detector temperature range dynamic switching method provided in the above embodiment.

[0159] As another aspect, the application further provides a computer-readable storage medium. The storage medium can be included in the detector control system described in the above embodiment, or can exist independently without being assembled into the detector control system. The storage medium carries one or more computer programs. When the one or more computer programs are executed by a processor of the detector control system, the detector control system implements the environment-adaptive detector temperature range dynamic switching method provided in the above embodiment.

[0160] The above-described embodiments are only used to illustrate the technical solutions of the present application, but not limit the present application; even though the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

[0161] In the above embodiments, according to the context, the term "when" can be interpreted as meaning "if" or "after" or "in response to determining" or "in response to detecting". Similarly, according to the context, the phrase "upon determining" or "if detecting (the stated condition or event)" can be interpreted as meaning "if determining" or "in response to determining" or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)".

Claims

1. A method for dynamically switching the temperature range of a detector based on environmental adaptation, characterized in that, The method is applied to a detector control system, and the method comprises: acquiring infrared thermal imaging data, identifying high-temperature pixel points in the infrared thermal imaging data that exceed a preset temperature threshold as a target high-temperature region; The step of acquiring infrared thermal imaging data and identifying high-temperature pixel points in the infrared thermal imaging data that exceed a preset temperature threshold as a target high-temperature region specifically comprises: based on historical posture coordinate data and historical infrared data collected by multiple associated detectors, a three-dimensional scene map containing multiple known heat source positions and known heat source characteristics is constructed; the associated detector is another detector that is located in the same monitoring area as the target detector and has the same device function; infrared thermal imaging data is acquired, and a high-temperature combined region of all high-temperature pixel points in the infrared thermal imaging data is extracted; the estimated position of the combined region in the three-dimensional scene map is calculated according to the image features and posture coordinate features of the high-temperature combined region; based on the estimated position, the known heat source positions and the known heat source characteristics, multiple known heat source regions and a new heat source region are determined as the target high-temperature region; The step of constructing a three-dimensional scene map containing multiple known heat source positions and known heat source characteristics based on historical posture coordinate data and historical infrared data collected by multiple associated detectors specifically comprises: reading historical posture coordinate data and historical infrared data collected by multiple associated detectors from a state record database, and constructing a data index table according to the collection device ID and the time stamp; according to the time stamp order in the data index table, the historical posture coordinate data is subjected to coordinate system conversion and registration to generate spatial point cloud data in a unified coordinate system with time sequence correlation; the historical infrared data is mapped to the corresponding spatial point cloud positions in the spatial point cloud data, and heat source characteristic parameters of multiple spatial points in the spatial point cloud data are calculated; according to the spatial point cloud data and the heat source characteristic parameters, a three-dimensional scene map containing known heat source positions and known heat source characteristics is constructed; The area proportion of the target high-temperature region in the full picture of the infrared thermal imaging data is calculated to obtain a high-temperature region proportion value; The area proportion threshold and the duration threshold corresponding to the current scene mode in the mode strategy library are determined; When the high-temperature region proportion value exceeds the area proportion threshold, a timer is started, and the high-temperature duration is determined based on the timer; When the high-temperature duration exceeds the duration threshold, the temperature change rate of the target high-temperature region is calculated according to the infrared thermal imaging data of multiple continuous frames in the high-temperature duration; When the temperature change rate exceeds a preset change threshold, an instruction is sent to the control circuit of the thermal imaging module, so that the target detector switches to a target measurement range to monitor the target high-temperature region.

2. The method of claim 1, wherein, After the step of determining multiple known heat source regions and a new heat source region as the target high-temperature region based on the estimated position, the known heat source positions and the known heat source characteristics, the method further comprises: acquiring a known identity of the known heat source region and binding a new identity to the new heat source region; Bind real-time temperature data, time stamp and collection device ID for the known identity and the new identity, generate data update record, and upload the data update record to the state record database.

3. The method of claim 1, wherein, After the step of reading a plurality of historical posture coordinate data and historical infrared data collected by the associated detector from the state record database, constructing a data index table according to the collection device ID and the time stamp, the method further comprises: Based on the data index table, grouping the historical infrared data of each collection device according to the monitoring area, calculating the spatial distribution density and the temperature cumulative duration of the infrared heat source in each monitoring area; According to the spatial distribution density and the temperature cumulative duration, identifying the heat source gathering point and the heat diffusion channel in the area, and constructing a heat source activity map; Superimposing the heat source activity map on the corresponding posture coordinate data in the data index table, establishing a mapping relationship between the heat source activity area and the spatial position; Based on the mapping relationship, classifying the spatial area in the three-dimensional scene map according to the heat source activity degree, and adjusting the sampling density and the monitoring frequency of the target detector in different areas according to the classification result.

4. The method of claim 1, wherein, Before the step of calculating the temperature change rate of the target high-temperature area according to the continuous multiple frames of infrared thermal imaging data in the high-temperature duration when the high-temperature duration exceeds the duration threshold, the method further comprises: Calculating the temperature average value of all pixel points in the current field of view of the detector as a temperature reference value; Traversing each pixel point in the current field of view, marking the pixel points with a temperature value higher than a preset multiple of the temperature reference value as high-temperature points; Performing connected component analysis on the high-temperature points, merging adjacent high-temperature points into standard high-temperature areas, and assigning a unique identifier to each standard high-temperature area; Recording the pixel number, area centroid coordinates and area boundary coordinates of each standard high-temperature area, and constructing a region attribute list.

5. The method of claim 4, wherein, After the step of recording the pixel number, area centroid coordinates and area boundary coordinates of each standard high-temperature area, and constructing a region attribute list, the method further comprises: Based on the region attribute list, extending the boundary coordinates of each standard high-temperature area, and adding a preset distance of early warning buffer zone outside the original boundary; Adding the pixel points of the early warning buffer zone to the region attribute list, expanding the region boundary coordinate record, and obtaining an extended region attribute list; Extracting the centroid coordinates of each standard high-temperature area after extension from the extended region attribute list, calculating the distance relationship between adjacent regions, and constructing a region distance index; According to the region distance index, identifying the region combination with a distance less than a preset distance threshold, and generating a merging warning information.

6. A detector control system characterized by, The detector control system comprises one or more processors and a memory; the memory is coupled with the one or more processors, the memory is used to store computer program code, the computer program code comprises computer instructions, and the one or more processors invoke the computer instructions to make the detector control system execute the method according to any one of claims 1-5.

7. A computer-readable storage medium comprising instructions, wherein: When the instructions are run on a detector control system, the detector control system is caused to perform the method of any one of claims 1-5.

8. A computer program product, characterised in that, When the computer program product is run on a detector control system, the detector control system is caused to perform the method of any one of claims 1-5.

Images